Is the NBN building a regional divide?

Is the NBN building a regional divide?

Summary: Was the deal between Telstra and the National Broadband Network Company (NBN Co) just another nail in the coffin for a broadband strategy that is going to miss out on its main objective?


Was the deal between Telstra and the National Broadband Network Company (NBN Co) just another nail in the coffin for a broadband strategy that is going to miss out on its main objective?

The government has always indicated that one of the key aims of the NBN is to ensure that users in regional Australia have the same level of access to those living in the capital cities. Simon Hackett, CEO of Internode, has argued a lot lately that having too many points of interconnect (PoIs) will mean that smaller providers will not be able to afford to hook up the NBN in regional locations, because there won't be enough customers to warrant the expense. That will leave residents with the choice of just one or two big players.

There's also the issue of backhaul. Whilst the Australian Competition and Consumer Commission (ACCC) recommended that PoIs should be located where there is a contestable backhaul market, is the existence of two providers enough to ensure the best wholesale price? Particularly if the assumption is now that the decommissioning of its copper network removes the need for the full separation of Telstra.

And what commercial terms have been agreed in the deal with Telstra? We know the NBN will provide last-mile access at the same price to everyone, but what are they buying from Telstra in the way of backhaul and other network services — and at what price?

Simon Hackett says that the full commercial terms of the agreement should be available for everyone to see. After all, it's taxpayer money being spent here.

So, the question is: does the ACCC need to reconsider its decision on the location of points of interconnect? Or is it too late? Has it all been ratified by the signing of the agreement with Telstra?

Also this week, my thoughts on Virgin Mobile's new "Fair Go" campaign. Why do mobile phone companies have such difficulty in making life simpler?

Topics: NBN, Broadband, Telcos, Telstra


Phil Dobbie has a wealth of radio and business experience. He started his career in commercial radio in the UK and, since coming to Australia in 1991, has held senior marketing and management roles with Telstra, OzEmail, the British Tourist Authority and other telecommunications, media, travel and advertising businesses.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Please correct me if I'm wrong, but aren't there now at least four companies offering competitive backhaul to the 121 Points of Interconnect: Telstra, Optus, AAPT and NextGen? I haven't seen an actual announcement from Optus to that effect, but I believe their fibre network is more than extensive enough to cover the POIs.

    It's been heartening to see companies like AAPT and NextGen jump into this bit of the market - it's exactly what was hoped for, and is probably only just the baseline level for more that is to come.

    There's been a bit of rumbling in recent about the quantity of dark fibre that has remained dark - by prodding the competitive juices, this can be put to good use, and offer absolutely loads of bandwidth. I think this is where the 121 POI plan wins over the original 14 POI plan put up by NBN Co - it gets well ahead of the curve in preparing for a network that has ample capacity as well as the benefits of a competitive market, and is well able to avoid congestion in domestic routes. To be sure, international capacity is another matter - but that problem is an entirely separate one, NBN or no NBN.
    • Phil the title of the article was building a regional divide. The 14 points or 121 points is an interesting argument but not fundamental to whether a regional divide is being created. From an engineering perspective, and if the NBN had followed a modern approach to its network design it would be a serious issue constraining what should essentially be a series of interlocking rings making use of existing telco assets.

      But that is all irrelevant - the bush is getting serviced by wireless and sattelite. If one looks at NBN Co's own preditions these will not be able to deliver the capacity needed in about three yesrs time. To make matters worse if these excluded towns try and install their own fibre they will now fall foul of the governments legislation. That is where the digital divide starts with its enevitable drain of rural youth to the cities and so on.

      I would love to see an article on where the fibre NBN is actually going. You need to bear in mind Quiglies comments that they will not break out of the blackspot fibre for small towns.
  • 14 POIs was always the technically optimal architecture, comprising two geographically separated interconnection points in each capital city. The ACCC decision to increase it by more than 100 extra facilities was a political response to lobbying by longhaul fibre owners, with no technical benefit, indeed it has a number of previously canvassed negatives for technical robustness and of course greater operational costs to NBNCo.

    So are 120 POIs better or worse for competition? NBNCo needs fibre backhaul and will therefore either build or lease it, so in a 14 POI world, longhaul owners still gain NBNCo as a client on top of their existing government and corporate VPN customers. But small RSPs without their own backhaul can more cheaply buy competitive capital city fibre links to service customers on 14 POIs than expensive longhaul to reach regional customers at a further hundred locations. Sure, aggregators will come in and sell multiple regional links to small providers, but this means a markup on the connection cost to the RSP and is therefore directly bad for competition. Only big providers with their own backhaul will get a customer for $24 wholesale, while small ones will have to pay a few dollars more.

    NBNCo has been brilliant at finding the best technical solution for cost-effective universal delivery of fast broadband. ACCC may have thought they had it right last November when they did this, but the above simple facts prove otherwise.

    14 POIs will guarantee the lowest wholesale cost base, common to all geographical locations, and longhaul owners will still have huge revenue streams from their assets, which are, of course, only in the most profitable routes anyway.
    • I know Simon Hackett made many of these points. And a case can be made for a number of POIs which is larger than 14 but smaller than 121. Fair enough - there will never be complete agreement over where the sweet spot is.

      But the role and purpose of the NBN is to cover the so-called "last mile" of fibre, the element which is least economical to duplicate and which represents the greatest barrier to rollout of fast broadband over a large area. (Yes, the "last mile" is often many times more than that - especially in regional areas, but the principle is the same.)

      If there is already multiple choice for high-capacity fibre to a POI location, then it is, by definition, not impractical and uneconomic to have duplication of fibre to that point. And where there is duplication, there is the potential for competition, and the benefits of competition.

      Put simply, an RSP will have to pay for domestic backhaul one way or another. The first option is to have NBN Co provide all of this except for the intercapital legs; the cost of that will then be passed onto the RSP in the wholesale pricing. The second option is to have competition over that non-intercapital leg, or have RSPs use their own resources, where they have them; in either case, there is the opportunity for the cost to be lower.

      Over time, the result is probably better this way. I was worried that there would be little impetus for competition over the short time, but this fear has been allayed with the entry of 4+ players already into the backhaul market.
  • (First words above should of course be "14 POIs" not "4" - had to copy-and-paste when the post operation failed, and I truncated the first 1.)
  • I think the biggest deciding factor will be how it's rolled out.
    If it is rolled out in batches of complete POI areas then we should be fine.
    But if it's rolled out at 120 locations simultaneously I can see the resellers having a fit trying to keep up.
    Batches of 15-30 areas would be ideal in my mind.
    Ie you pick 15/30 areas and you roll out to 100% of homes in that area completing the entire area before moving on to the next area.
    Do that 4-8 times and you have completed the network.
    You've also reduced the load on resellers in getting their own network arrangements going. Rather than a sudden hit of 120 locations, it's 15-30 every 12-24 months.
    Should stop baldness in network admins.