An NBN restrospective - Part 2.

Having now fixed nbn's culture problem (and re-rebranded it to NBNCo), we're ready to get into the minutiae of actually fixing this travesty of a network. We're going to make some key assumptions that will make this process a lot less painful.

The NBN is put on budget as a national infrastructure expense (it is currently ostensibly paid for with government bonds)
The Statement of Expectations has been altered to remove the profit motive (at least for now)

Basically, we're assuming that money is no object. With the NBN now on the Government's budget, it can direct funds directly to the project. This is currently not possible, and NBNCo has to ask its shareholders (currently Mattias Cormann and Paul Fletcher) for cash injections. This doesn't work when the Government's core election promise was to achieve a budget surplus, and doubly doesn't work when they don't actually care about the NBN.

Altering the SoE allows NBNCo to just plough money into the network with no expectation of making a return on it. Some good the expectation of a return is doing now anyway, the company only recently posted a record $2bn loss. Good news is that this doesn't have to be permanent, and as part of the technical solution for fixing this steaming pile, we will also be looking at how the NBN makes money.

Considering the needless complexity of the current rollout, which precludes a simple catch-all solution, we will be going through each current rollout technology individually. We will briefly discuss each technology, why it sucks, and what can be done to make it not suck.

By now, everyone in the country should be aware of what a mess Fibre to the Node (FTTN) is. Originally touted by Tony Abbott and Malcolm Turnbull to be a cheaper, as-good alternative to Labor's Fibre to the Premises (FTTP) network, the technology has been nothing short of a spectacular failure. Such was the magnitude of its failure that very shortly into its rollout, NBNCo (and PM Turnbull) were forced to drastically change tack, and began casing out the possibility of eating up Telstra's and Optus's old Cable networks.

Dr Ziggy Switkowski, the current NBNCo chair, fronted a 2003 Senate inquiry into the state of broadband access in Australia. It was at this inquiry that he declared Telstra's phone line network to be, "Five minutes to midnight," basically dead in the water (sometimes literally). It is this same Dr Ziggy Switkowski who was charged by the Liberal Party to roll out an FTTN network utilising this same Telstra phone line network.

FTTN is really only suitable for short lengths of copper. Fibre is run to a box somewhere in the neighbourhood, in which DSL line cards are installed and connected to the various phone lines running to houses. The flavour of DSL used in FTTN drops in speed precipitously with distance from the node, performing around the same as traditional ADSL 2+ at around 1km of line length from the node, assuming the line is in good nick.

NBNCo found out very quickly that Dr Switkowski was not lying for a change in 2003 and many, many phone lines in this country simply not suitable for upgrading to FTTN. Regardless, the Government made the company forge ahead until complaints of slow speeds and dropouts became so deafeningly loud that they could no longer ignore FTTN's shortcomings. Today, the FTTN footprint covers about 30% of the population, a far cry from the original plan of around 76% coverage.

Around 15,000 premises in the FTTN footprint are unable to achieve even 25Mbps download, the speed that was promised to be guaranteed by 2016, and the minimum speed in the Government's own SoE. The ACCC recently came down like a ton of bricks on both NBNCo and the ISPs for failing to identify customers who were paying for services they could not achieve on FTTN. Customers were being charged for a 50Mbps service while barely receiving half, or even a quarter of that in some cases.

There's no two ways around it. The entire FTTN rollout was, and remains, a total failure. There is only one logical solution, and that is to completely tear it out and replace it with FTTP. Given NBNCo's SoE mandates them to provide a baseline of 25Mbps to all premises in the country, we would naturally start with those 15,000 locations unable to attain 25Mbps downstream. NBNCo have access to the line condition of every single FTTN connection in the country, and as such once the priority sub-25Mbps connections are fully remediated, they can assign crews to overbuild the FTTN in order of line quality from worst to most acceptable.

FTTP, due to its use of fibre optics all the way into the home, guarantees that the end user receives the speed they pay for, be it 25Mbps or 2500Mbps. FTTP is also future proof; higher speeds can be unlocked simply by replacing the electronics on each end of the fibre, and current fibre optic technology is capable of delivering a throughput of around 27Tbps (yes, terabits per second) over a single connection.

Most importantly however, FTTP is by far the most reliable communications technology available. Unaffected by the weather, power outages or even simply your neighbour's cheap solar inverter (believe it or not, solar inverters have been responsible for many DSL interference problems), FTTP can remain active pretty much indefinitely with no degradation of service, a welcome change for FTTN users stuck on drop-out prone lines.

With FTTN out of the picture, NBNCo actually has a shot at delivering on its Statement of Expectations, as well as delivering quality telecommunications to a segment of the population which has been totally neglected for the past seven years. With access to the Internet being viewed as a basic utility by most now, it is a disservice to the nation to leave people in the FTTN footprint high and dry with last-century broadband speeds and developing nation electricity grid levels of reliability.

Continue reading in Part 3.

Notes from 2022

An NBN restrospective - Part 2.