HFC agitators ignoring BSS hit

It is astounding how many in the telecoms sector desperately want to assign value to a network that has only ever been a marginal player in Australia's market. Like flat-earthers or climate-change deniers, they insist that things are a certain way, even though evidence and experience suggests otherwise.

Enter Michael Porter, a Deakin University academic who has developed a new plan for a Coalition Government to reduce the cost of the National Broadband Network, save Australia’s broadband customers from the pain of having a modern communications network, and reverse the Australian Competition and Consumer Commission's approval of NBN Co's AU$800m deal to buy Optus out the hybrid fibre coaxial (HFC) market.

It was always going to be a contentious deal, representing what is in textbook terms a blatantly anti-competitive move and causing Malcolm Turnbull fits of press-release apoplexy. The contract has, of course, been warmly welcomed by Optus, which sees the deal as a great way to recoup its investment in what has been, at best, an underperforming fixed-network asset — and, at worst, an albatross around the company’s proverbial neck.

Diversity is the enemy of the BSS and its companion OSS, which lies at the heart of any telco network, and ensures that services run smoothly and customers get billed correctly.

Before you go ranting about the need for fixed-infrastructure competition, remember that Optus' network has been severely aggrandised in the minds of those arguing for its preservation. Unlike Telstra's relatively more-extensive HFC network, the Optus cable only services are, according to Optus itself, available to "selected homes in Sydney, Melbourne and Brisbane. Service is not to be used for commercial purposes".

There are two key points here: firstly, that this network is a paper tiger; it's only available in a few parts of Australia's three largest cities, so ascribing it value as some sort of ubiquitous broadband competitor is blindly optimistic. Its 429,000 subscribers represent less than 5 per cent of Australia's premises.

The second key point is that the service is not to be used for commercial purposes. This restriction, of course, is there so that companies won't flood the network with large volumes of data that Optus would prefer were shunted over more technically-capable networks; as I, and every other Optus cable user know, HFC performance can vary dramatically, depending on hourly usage volumes and the introduction of new services, like Optus MeTV, tends to impact performance for everyone.

Another issue that proponents of HFC preservation fail to understand is that perfectly competitive HFC networks are beneficial to nobody. In the US, cable networks have so many markets to choose from that providers pick and chose the most profitable areas to roll out infrastructure, where they can effectively cement infrastructure monopolies.

Look at the coverage map for cable provider Cox Communications, for example, and you'll see a highly fragmented network that cherry-picks many of the most populous areas. Some areas may have overlap between different service providers, but, on the whole cable, infrastructure is an opportunistic investment.

It's only because of that country's mature market that most homes can get cable — but Australia's population is so heavily weighted towards the city that it's clear the private sector will never follow the same path. The private sector was only ever interested in dense capital-city areas — and on this point, Australia blew it: while the desire for competition drove two mirrored HFC roll-outs throughout some of Australia's cities, infrastructure competition was simply untenable. That's why Optus stopped trying to compete with Telstra's roll-out, and it's why Optus is desperately eager to get rid of its HFC network.

Simply put, HFC is an artefact of 1990s-era competition idealism, not a viable part of a 21st-century broadband network. So while academic models might enable the preservation of the Optus HFC network — and counteract what Porter calls "a travesty of economics [that] undermines a competitive pro-investment strategy" — they have little resonance in reality.

Where, after all, is this "pro-investment strategy"? Optus hasn't invested in its cable network for over a decade. This seems to be a particularly dramatic example of economic theory overriding telecoms reality.

Yet, the biggest surprise is in Porter's alternative suggestion: that the NBN be shifted to an area-by-area roll-out model that would see broadband tenders awarded based on the technologies already available in a given area.

Porter's model would create a patchwork of varying technologies and competing commercial interests, which would only perpetuate geographical service discrepancies.

This is another of those ideas that sounds good in theory, but would be a disaster in practice. While Porter's model might allow the staged targeting of the most broadband-deficient areas, it would create a patchwork of varying technologies and competing commercial interests, which would only perpetuate geographical service discrepancies — and reassert Telstra's monopoly in rural areas, whose low population would make them commercially unattractive to competing tenderers.

Not only would this approach muddy the waters of service consistency, but it would create an operational nightmare for any authority trying to manage this service patchwork. Diversity is the enemy of the business support system (BSS) and its companion operational support system (OSS), which lies at the heart of any telco network, and ensures that services run smoothly and customers get billed correctly.

BSS/OSS investments are always significant and usually incredibly complex: even NBN Co, which is building an architecturally homogeneous network, spent AU$124 million on its BSS/OSS and related components — in the last six months of 2011 alone — to ensure that it can provision and support its services around the country, in the same way for all subscribers.

Time has repeatedly shown that diversity is the enemy of the BSS: once you have different networks to support, you have to service customers on those networks using different back-end systems, which have inevitably been designed and implemented with different parameters and service levels.

Ironing out the wrinkles in such systems, which usually requires exhaustive and exhausting systems integration or migration, is a major reason why telecoms mergers have struggled in the past. It's the reason competitive ADSL providers have found it hard to match Telstra's customer service levels: because they historically lacked visibility into its network and, therefore, customers had to take its word that it was doing anything for their customers, at all. And, in many cases, it wasn't.

While an area-by-area roll-out might achieve the policy goal of capitalising upon existing infrastructure, it would create an NBN that was operationally dysfunctional and, in all likelihood, incredibly difficult to upgrade for the future. It would disincentivise private investment, perpetuate the pained existence of small networks, with little viability on their own, and ensure that our broadband future remained tied to the scatterings and smatterings of 15 years of opportunistic infrastructure investment. Forgive my scepticism, but that doesn't sound like a path to anything but chaos.

What do you think? Does area-by-area tendering promise a better NBN? Is it just the latest attempt to denounce Labor's NBN plan? Or should we be preserving existing HFC assets at all costs?