NBN Co, the company building Australia's National Broadband Network (NBN), has announced that from 2017 it will offer DOCSIS 3.1 over the hybrid fibre-coax (HFC) networks it acquired from Telstra and Optus.
The announcement is surprising because just two-and-a-half weeks ago, on February 23rd, NBN Co spoke of designing its network for future DOCSIS 3.1 deployment, but didn't hint it was planning to do so any time soon. It's now saying, in a statement titled "NBN Co to unleash fibre speeds for cable customers" and saying “A new technology standard that promises to deliver gigabit speeds is planned to be rolled out across Australia from 2017, NBN Co announced today.”
That new technology is DOCSIS 3.1, a product of the US pay TV industry’s Cable Labs research outfit. DOCSIS promises broadband connections at up to 10Gbps down and 1Gbps up, over existing cables after core network and node upgrades. Interoperability tests have reached an advanced state: Cable Labs says commercial DOCSIS 3.1 product should be available in mid-2015. DOCSIS 3.1 is about ten times faster than the current standard, DOCSIS 3.0.
NBN Co chief technology officer Dennis Steiger announced the plan to adopt DOCSIS 3.1 at an event called “Cable Congress 2015” in Brussels.
Steiger also says, in the statement we've been sent, that the decision to adopt DOCSIS 3.1 represents “... the same philosophy that’s driving our entire multi-technology rollout.”
“All the technologies we’re using have an upgrade path to deliver higher speeds and greater capacity,” he added.
And indeed they do: fibre-to-the-premises has plenty of carrying capacity waiting for new head-end kit to unlock. Vectored DSL's successor, G.fast, was signed off last December and chipsets to turn the standard into products are close to going on sale.
But NBN Co is yet to announce a G.fast roadmap.
+Analysis The mention of "fibre speeds for cable customers" suggests that NBN Co is acknowledging community concerns that its multi-technology-mix plan won't deliver the headline speeds promised by the previous government.
Communications Minister Malcolm Turnbull will therefore, we expect, make much of the decision to adopt DOCSIS 3.1, as it supports his argument that his administration is finding a way to deliver scalable broadband infrastructure cheaper and faster than the previous government's exclusive fibre-to-the-premises plan.
News of the DOCSIS 3.1 adoption also explains Turnbull's statement to Parliament last week that he sees no need to future-proof the NBN”.
In light of the DOCSIS 3.1 announcement his words now take on different meanings. “What does the customer need?” he asked. “Very good broadband. What technology should be used? Whatever technology is most cost-effective at the time.”
DOCSIS 3.1 clearly meets those criteria.
Does G.fast? Perhaps not today: the main backer of the standard is the upstart Sckipio, with better-known chipmakers signalling adoption later this year. But before long G.fast will be as widely-accepted as DOCSIS 3.1.
If NBNCo soon announces it will adopt the second standard too, Turnbull may have reason to feel his victory over the previous government's model is complete because he'll be able to show he's delivering 100Mbps connections without the alleged disruption and cost of the FTTP NBN plan.
Without a G.fast announcement, there will be arguments to be made about equity and commercial imposts. The equity argument is simple: why should Australians be offered services with different potential? Commercial objections would lie on the long-held hope that the advent of near-universal broadband in Australia would spur local innovators to create new products and services. If the NBN forks into 100Mbps-capable and 10Gbps-capable branches, what's a developer to do? Develop for the lesser service and the faster offering? Or go for the lowest common denominator?
Industry won't welcome that choice. I'm betting Turnbull will be trying to make sure it doesn't need to by moving NBN Co towards G.fast adoption. ®
Bootnote: We told you DOCSIS 3.1 would happen, back in 2013.