Sunday, September 30, 2012

Independent Australia: pro-NBN works better with accuracy

A period of some hours over at Independent Australia isn't sufficient to get a comment moderated, so my comment is here.

I'm pro-NBN, but I like to defend it with facts. So while it's good to see Independent Australia chiming in on the debate, here, there are some points I'd like to correct.

1. "Turnbull’s go to line that “[NBN Co] has only had about 5,000 connected”. Having studied the business plan, I see nowhere that Turnbull could possibly have gotten this figure from".

It didn't come from the business plan. Mike Quigley used a similar number here.

"In terms of activations, we now have close to 5½ thousand premises connected to the NBN across the country."

This was in February and is presumably obsolete, and anyhow included satellite (which is even worse for the fibre case).

My guess is that a statement like this has become part of Turnbull's armoury - even if it is now out-of-date.

2. "...the bandwidth falloff for VDSL2 Profile 8 (more than likely what will be deployed to keep costs down & deliver service to 1kilometre) not even making it to 80Mbps at the DSLAM (Digital Subscriber Line Access Multiplexer, i.e. Exchange Equipment)."

That probably needs a re-write. The DSLAM is where the bandwidth is delivered from. So it's hard to discern the point that's being made.

3. "...a minimum of two pairs (bonded) are required to run the service, and one pair to run a telephone line." (my emphasis)

The phone line runs on one of the pairs, just as it can with ADSL: the lowest frequencies are reserved for voice (described as USO services in the standards documents). The broadband signal starts above 100 kHz.

4. "The requirements for lead-ins is that two pairs be installed to the first socket in the premises, the Telstra/Optus networks are both designed with this in mind, leaving little room for two pairs to be taken up by a VDSL2 service."

Optus doesn't run a twisted pair copper network. Its broadband services are delivered over HFC, or over Telstra copper - as are its fixed line voice services.

5. "there would have to be, at the least, a new copper pair for every premises connected to the network."

That depends. Many houses have two pairs installed, not because they were needed: merely because years ago, that's what Telstra did, just in case one failed. What we don't know - among the many things we don't know about the copper network - is how many houses have two usable pairs installed.

So what's the point to me correcting this?

The NBN is a political debate - which means that technical errors get exploited to make a political point. It's not a good idea to put weapons in the hands of your opponent.

1 comment:

Sortius Sortius said...

1. As I stated, Turnbull is taking his number from the projected connections, which were surpassed way before June 30th, as Turnbull claims they are from.

2. With xDSL being determined by distance, this is quite relevant when, as he stated in the 730 interview, 1/3 of connections will be 80Mbps.

3. Profile 8 (which is what will more than likely used, read the whitepaper I linked) requires bonding (utilising 2 pairs, not 1) to get anything near 80Mbps to 1/3 of subscribers.

4. HFC is unacceptable as a service. Unlike xDSL & fibre optics, the bandwidth is shared among users. So 100Mbps HFC, shared among even 50 users (normal use is 500-2500) brings the shared speed to less than 2Mbps if you take into account overheads. Add to this that HFC is at the limit of possible speed development for coaxial cable, fibre is not.

5. You're essentially agreeing with my premise that VDSL isn't going to work in Australia. Basic VDSL2 takes over the whole phone line (pair), so either VOIP has to be supplied, or the 2nd POTS line made active.

I'm not sure why you didn't mention me in your response, I'm hazarding a guess so I couldn't respond. Your points. I try to explain the technical side of these arguments as best I can, supplying as much technical data as possible. If you'd taken the time to read the whitepaper & the NBN Co business plan you may have understood my article better.