An NBN retrospective - Part 1 ? X
Home GitHub Contact

Notes from 2022

I wrote this in February of 2020. I have had extremely strong opinions on the NBN for most of my life, and I'm still mad about what happened to it. This was originally published in 'The Queenslander' - an online magazine which has since gone defunct. You'll have to forgive any grammatical errors as this text was extracted from an old version of the document I submitted to the publication before editing.

Oh, and a note for the non-Australian: the Liberal Party is Australia's "mainstream" right wing party and bears no resemblance to similarly named parties in other countries. In more recent years, they have been falling under the increasing influence of outwardly illiberal far right actors seeking to emulate the United States Republican Party.

The original article (well, really it's more like a report) was over 30 pages long, so I'm splitting it into parts to make it more easily digestible.

This is Part 1, which focuses primarily on cultural problems at NBNCo brought about by changes made in the wake of the Liberal Party's 2013 election victory.

Part 2 focuses on the failings of FTTN.

Part 3 deals with the HFC network and ways in which it can be salvaged.

Part 4 explains why the LTE network was poorly thought out from the start.

Part 5 briefly touches on the economics of the NBN.

Part 6 hammers home the fact that all of this is a result of shameless and disgusting short sighted political manoeuvring by the Liberal Party to protect the interests of Rupert Murdoch and other corporate friends such as Telstra.

An NBN retrospective - Part 1.

It's finally 2020, and as the NBN rollout starts to wind down now is a good time to take stock of just what the flipping heck happened over the last 10 years. We were promised in 2009 that by the end of this year we would have an all-fibre access network to replace Telstra's and Optus's dilapidated twisted pair and HFC access networks. Instead, thanks to some absurdly transparent political manoeuvring around 2013, what we actually got was...

Telstra's dilapidated twisted pair and HFC access networks under new management. And while this was good for my Telstra shares, it was and still is an absolutely diabolical, unforgivable, politically-charged sabotage of objectively good national infrastructure, intended to do nothing more than prop up Rupert Murdoch's crumbling media empire and to Fortnite dance and then piss on the political legacies of Kevin Rudd and Stephen Conroy.

Since this change in strategy, even mentioning the NBN in less technical circles gets nothing but an exasperated groan and tales of incessant dropouts, slow peak hour speeds, and a government agency intent on flatly denying that there's a problem. This, as has already been discussed, is by design. But now that the rollout is finished, what can be done to actually fix the problem?

Opacity

The biggest problem with the NBN right now has nothing to do with technology. With the change of government in 2013 came a new board for NBNCo, one that the company's two shareholders (being the Minister for Finance and the Minister for Telecommunications) knew would carry out their will without question. Gone was Dr Mike Quigley and his corporate ethos of transparency and accountability and in came Ziggy Switkowski, an ex-Telstra executive with ties to the government.

According to Switkowski and his new board of ex-Telstra executives, the first and foremost priority for NBNCo was to immediately redact all commercial documents pertaining to the rollout. From September 9 2013 until mid 2014, after the company's "independent strategic review" was completed, no internal document was available for viewing. When they became available again, most details that interested parties had become accustomed to having access were redacted and deemed to be "commercial in confidence" - information such as fault incidences, revenues and services connected.

This included most of the strategic review itself, including its frames of reference and key assumptions made when coming to the decisions it recommended (which conveniently enough for the government involved doing exactly what their election manifesto promised). During this time, all construction work was suspended, and an undisclosed sum was spent on rebranding the company with all references to NBNCo being replaced with nbn(tm). A worthy use of taxpayer funds from the Strong Economic Managers.

By the time the FTTN/FTTB/FTTC/HFC/anything-that-isn't-FTTP rollout was underway, this corporate culture had become entrenched. Even nbn's customers, the ISPs, were left in the dark about most operational matters to the point where they were entirely powerless to deal with end user complains about slow speeds and dropouts. ISPs would refer customers to nbn who would deny any problem on their end and flick them back to the ISP. This is a problem that persists even today, despite the ACCC's intervention forcing RSPs and nbn to publish speed data and refund customers unable to achieve the speed tier they're paying for.

In virtually every HFC deployment around the world users are allowed to access their modem's statistics page, which gives extremely important information about the quality of the signal entering the home. On the NBN, access to this page is disabled as soon as the CM8200 gets its configuration parameters from the remote end. There is absolutely no reason to do this other than to obfuscate problems with the HFC network, problems which the company denied for 3 years until they became so widespread and so disruptive that they were forced to put the entire rollout on hold to deal with them.

This "pause" as they called it took the better part of a year and still problems persist in certain areas on the HFC network. Had these network metrics been available for public viewing (as they should be) the company would have been forced to act much sooner. People only noticed because Telstra Cable and Foxtel share the same network infrastructure, so those services were also impacted by nbn's activities.

Transparency is a necessary facet of accountability and right now nbn are doing all that they legally can to avoid being accountable to anyone. The simple fact is that operational transparency highlights the great flaws in what has been deployed when compared with what was meant to be deployed.

The problem with this pathetic attempt to save the Liberal Party's face using taxpayer money is that it leaves consumers and ISPs totally powerless to resolve disputes with nbn unless nbn deems the dispute to be "reasonable". Currently, the criteria set by nbn for a "faulty" FTTN service is either a downstream sync speed lower than 20Mbps or 16 (yes, sixteen) dropouts per day with no more than 1 hour between dropouts.

If a service syncs at 21Mbps or only drops out 15 times a day, it is considered to be within spec by nbn, regardless of whether or not there is actually an issue with the connection. The company's contempt for the end user and their own customers is on full display here. The entire corporate structure exists not to facilitate the rollout of a National Broadband Network, but to actively frustrate that process.

The only way to get the NBN truly back on track is to simply demolish the entire corporate structure and return to a positive culture of operational transparency and accountability, free from the influence of ex-Telstra corporate stooges with deep links to high profile members of any particular political party. Like any government body, nbn should be accountable not only to its two shareholders but to everyone in the nation as it used to be. The NBN is being run in a manner similar to Home Affairs and ASIO which is quite simply disgusting for a national infrastructure project.

Productivity (but only for me and my mates)

Public-private partnerships are a spook. And in the case of the NBN, they're a bigger spook than the bogeyman. There's not a lot to say here except for that the NBN is being built by layer after layer after layer of contractors, each one taking their cut off the top before passing down what's left until the guy actually splicing the fibre on the street is on barely above minimum wage. This was a decision made in the very early days of the rollout, as it was decided that amassing and training a workforce would be too costly and take too long. This decision, however, was a poor one.

The only reason the private sector was so quick to get going was quite simply because they cheated. There are people working on the HFC network right now who failed to demonstrate the basic competency of crimping an F connector yet were given their ticket to perform work by dodgy RTOs in exchange for a case of beer. These same RTOs also give their students the answers to the theoretical components involved in becoming certified, leading to a fundamental misunderstanding of standard practice or how various network elements are meant to behave.

The contractors themselves are complicit in this workforce deficiency, too. In order to cut costs (not to save the Government money, but to increase profit margins), most contractors exclusively subcontract to 457 visa holders. There have been instances of fully qualified Australian citizens being fired simply to make way for temporary work visa holders. There has been at least one instance of this occurring because the citizen refused to ignore workplace health and safety regulations where his colleague did not.

The NBN was meant to see Australia return to a publicly owned last-mile access network, as we had until Telstra was privatised. Before said privatisation, Telecom Australia had an entirely in-house workforce of highly trained professionals building and maintaining their network. Such was the quality of this workforce that industry giants such as Ericsson, Siemens and Alcatel would frequently visit Telecom Australia's various training facilities around the country and take our practices back to Europe to share with the world. Telecom Australia literally set new world standards in telecommunications practice. nbn on the other hand is barely capable of meeting existing ones. The company has virtually no quality control measures in place and is entirely at the mercy of cowboy contractors just looking for a free ride on a government gravy train.

Now that the rollout is ostensibly complete, and as the various contracts nbn has with its contractors start to lapse, the company must shift towards an internally trained workforce of linesmen, engineers and technicians in order to properly service (and upgrade) this monstrosity they have built with at least a modicum of quality control. Being a telecommunications technician was once an honourable and respected career path for young men, and Telecom Australia was one of the country's most enjoyable companies to work for. There is absolutely nothing stopping nbn from carrying forward Telecom's legacy save for a boardroom and government entirely hellbent on transferring wealth from the taxpayer to their mates.

High standards (for 1995)

During Dr Quigley's tenure at the top of the NBNCo corporate structure, the company actually did look something like a true heir to Telecom's legacy. Hundreds of millions was invested in research and development, and from this came new standards for large-scale FTTP rollouts. NBNCo was instrumental in popularising the adoption of "skinny" fibre, ribbon-like fibre optic cables that are both cheaper to manufacture and easier to deploy in the field. NBNCo also innovated on removing the need for a Fibre Distribution Hub, instead replacing it with a more compact multiport joint closure which fits inside a standard double-length Telstra pit.

By April of 2013, the average cost per premises of deploying FTTP had dropped from the initial $2,400 during the rollout's infancy to a meagre $1,100 in brownfields - already existing homes. Interestingly, as soon as the government changed this figure seemed to increase by precisely 4 times to conveniently make it more expensive than the average cost per premises of FTTN (which now costs over $5,000 per premises in some cases).

With the change of rollout strategy, no longer was it necessary to continually innovate or spend money on creating new standards. Telstra had all the standards handily ready to go. In the case of the HFC network, such is the legacy of Telstra's standards that the brand new tags used by nbn to label customer drop cables nd outside plant refer to Telecom Australia with the orange and blue T, branding which has been defunct since 1995.

nbn were also happy to spruik the latest innovation in HFC technology, DOCSIS 3.1. Despite DOCSIS 3.1 being objectively good, it will never live up to the company's expectations. Since remedial work on the HFC network started, the company has been deploying ARRIS 1GHz and 1.2GHz rated equipment in the street - a requirement for full DOCSIS 3.1 compliance. However, inside those countless millions of beige nbn boxes and grey Telstra boxes on the sides of homes lives one of the many weak links in the HFC deployment - the isolator.

When Telstra first rolled out the HFC network, the original DOCSIS specifications used only a small portion of the available bandwidth on a coaxial cable, and as such built the network to a lowest common denominator specification of 750MHz. And despite nbn's insistence on activating DOCSIS 3.1, they continue to use the Telstra-prescribed 750MHz isolators in almost all brand new installations simply because this was Telstra's accepted practice. While DOCSIS 3.1 will work across these isolators, nbn will find themselves running into problems as they try increase bandwidth across the network past 750MHz for "Full Duplex" operation. The most insulting part is that Telstra's HFC network was second-rate even in the 90s compared to most of the world, so nbn are relying on decades-old standards that weren't even standard in their heyday.

As if "adapting" these standards wasn't bad enough, how about outright plagiarising them? When shifting to a mix of technologies, it was necessary for nbn to create documents outlining and explaining the various technical intricacies of this new bastard network. One such document is the Authority to Alter, or A2A, which details what parts of the network on a customer's property registered cablers are and are not allowed to mess with. Every version of this document, save for the most recent, was a simple copy-paste of Telstra's old A2A documents, most of which date back to the early-mid 2000s, when Telstra stopped deploying new network assets. In the grand scheme of things this is pretty minor, but it illustrates just how determined nbn are to be as wholly unremarkable and unambitious as they possibly can.

nbn's entire corporate structure is arguably the biggest problem with the rollout post-2013. A government enterprise sworn to secrecy by its shareholders to obfuscate the utter inadequacy of its "Cheaper, Faster, Better" network. A government enterprise hijacked by the Liberal Party and heir Telstra cronies to transfer wealth from the taxpayer to their private contractor friends. A government enterprise so blatant in their goal to simply continue on with Telstra's mediocrity that they go so far as to reuse and outright steal Telstra materials and documents from 20 years ago.

Sure, there are technical problems with the NBN that we will to, but these all stem from the deficient corporate culture imposed on the project by the Liberal Party. Even with the multitude of technical failings wrought upon the nation by the socioeconomic vandals in the Liberal caucus, by putting an end to the absolute rort that is the public-private partnership, increasing internal transparency and resuming the in-house development of work standards and methods, the NBN could slowly and steadily become something that Australia can actually take pride in.

Continue reading in Part 2.