Reconsidering the Coalition's NBN promise

Two details in the NBN rollout could potentially force the Coalition's "Better, cheaper, sooner" NBN to conform with the existing standards and roll-out plan:

• The NBN Access standard for a layer 2 VLAN-in-VLAN ethernet bitstream, and
• The deployment of Points of Interconnect (PoI).

NBN Co has moved into high-gear in deploying the core of the NBN access network: the PoI's and the supporting Transit Networks linking them back to customers. They plan to be mostly finished years by the end of 2014.

While there are seven active PoI's today, by the end of June 2013, around half of the final 121, agreed by the Australian Competition and Consumer Competition (ACCC), will be up and running and work on the rest will be well advanced.

POI considerations 

By the time the federal election comes along, NBN Co will be too far advanced in deploying PoI's to do anything but be allowed to finish the job. At the very least, the regulator will need to see a pretty convincing case from a Coalition government to consider varying the number of PoI's and/or their location. Considering how long it took to access and agree the list the first time, it's unlikely that a review could agree variations before the existing rollout plan was finished in 2014.

The underlying issue here is the fact that there is a sound basis for the selection of the PoIs and the technical implications of coalescing the last, say 25-40, PoI's into just a few are large.

Now each PoI won't be exactly identical, but they will be very similar and the backhaul 'pipes' for ISP's and the volume of traffic will be similar. Introducing a few, very non-standard PoI's, and not in major metro areas, will increase capital costs and operating costs for RSPs/ISPs. It's an idea that the telcos and ISPs will resist tooth and nail.

So it's quite likely that the Coalition will have to allow NBN Co to finish its PoI rollout.

The case for standardisation 

Then there's the small matter of "standard gauge" for customer access: the layer 2 ethernet bitstream.

The cornerstone of the NBN is "open access customer access network", which requires a single, standard means for RSPs/ISPs to connect with it. A process that potentially undermines this standardisation will be a hard sell for the Coalition. The standard exists for a good reason because we need to avoid the current confusion, similar to the multiple incompatible railway gauges built around Australia in the 1890's.

Infrastructure standards decisions live for a very long time. Australia isn't going to change electricity voltage or frequency anytime soon, nor will we ever do the far easier change, drive on the right side of the road.

The Coalition will have to mandate that its proposed "cheaper, sooner" local access networks comply with the existing NBN layer 2 standard and connect to the existing regulator approved PoI's.

We shouldn't rule out the Coalition's ability to pull a rabbit out of a hat and actually build a cheaper mixed media network: Fibre for the privileged, 12Mbps ADSL, HFC and perhaps WiFi or WiMax for the second- and third-class citizens.

But this mixed media network will have to conform to the existing Network Plan based around 121 PoI's and support the current layer 2 ethernet bitstream standard. It can't be a free-for-all or the whole thing reverts to the expensive mess that our rail system  turned into. This means creating versions of the standard client premises interface and the ONT (Optical Network Terminator), for each media type. That's an expensive hardware development project.

Given the history and the countless decades to partially rectify the confusion of railway gauges, the Coalition needs to be careful in not replicating the same mess on the information superhighway. 

This is an edited version of a blog post originally published on October 16. Steve Jenkin has spent 40 years in ICT in wide variety roles including large and small software projects, 7 years writing real-time Exchange software in a Telco and Admin, Software and Database work on PC's Unix/Open Source software and mainframes.