Over the past few weeks there has been a flurry of media noise around a proposal from the so-called G9 (a Singtel Optus led consortium) regarding Fibre to the Node (FTTN). The G9 have produced a whole lot of documentation but the details therein on the technology are particularly scarce (the academic in me can’t help ‘marking’ their attempt, which I rate as a definite ‘fail’).

However, one issue which they have raised is a proposal, originating from the ACCC in support of the G9’s scheme, for so-called “sub-loop unbundling”. I suspect that most people, even those with a technological bent, will have difficulty understanding what this is, let alone its implications, so let me try and explain (and in an anticipation of the response to comments I am sure I shall receive, in the interests of brevity, I have omitted a lot from this description).

The copper loop connecting the customer to the exchange progresses via a number of stages. Starting at the exchange it goes out via large, gas-filled, multi-pair cables (known as ‘main cables’) to the pillars, which are those objects that look like vertical torpedos that you see in the streets all over Australia. The pillar has a ‘distribution frame’ in it which enables the pairs in the main cable to be jumpered across to the pairs in the distribution cable. The distribution cables are smaller cables which go down the streets and drop off pairs into the ‘pits’ outside our homes. In the pit, a ‘lead-in’ cable is joined to the distribution cable and takes the signal into the home.

In today’s world, the equipment that connects the copper to the network – the multiplexors – resides primarily in the exchange. So if a customer churns to a competitive carrier, the concatenated pairs of copper cables that connect that customer to the network is disconnected from the Telstra multiplexor at the exchange and jumpered over to the competitor’s multiplexor.

Despite the fact that you may think that each pair of copper wires is a separate entity, you cannot simply connect any equipment to the wires because there is ‘cross talk’ between copper pairs in the same cable binders. This is analogous to being in an open plan office with partitions between cubicles. Even though you cannot see your neighbours, a bit of your conversation spills over into neighbouring cubicles. There are thus rules of the road for what you can connect to the copper wires, set by the Communications Alliance which is an industry body set up to establish the technical rules for the use of the network.

With sub-loop unbundling, the proposal is that the competitors can go and set up equipment at the pillars and take control of the pair of wires to the customer’s home from there. In theory they could go to any connection point in the network and do this. What’s the problem with this? Well, actually quite a lot. For a starter, there is a big difference between changing jumpers in the exchange versus changing them in the field. Believe it or not, the streets of Australia are a very hostile place – at least if you are a copper wire trying to connect a high speed electrical signal. Each time something in the ‘outside plant’ is touched you increase the risk of something going wrong. Our workforce have years of experience in working on this outside plant, but the risk would be much higher if someone inexperienced took it on.

However, an even bigger issue is the impact it will have on our ability to deliver high speed broadband to your home. If you have 2 homes served off copper pairs in the same binder group and one is served from the exchange and the other from a pillar, it becomes very difficult to manage the cross talk. Effectively either the customer served from the exchange or the customer served from the node gets a degraded service compared to what ought to be achieved. Think of that open plan office – if you are on the phone and your neighbour stands up and starts a conversation with someone at the other end of the office he is going to drown you out. And you speak louder thus ensuring that he can’t hear the person he is speaking to.

If this wasn’t bad enough, sub-loop unbundling also has another unintended consequence – it effectively drives the copper access network into a technological cul de sac. The speed with which you can deliver services over the copper keeps going up. You may recall that we started with modems at 300bps and today we are up to ADSL2+ with speeds up to 24Mbps (that’s approximately an 80,000 fold increase in about 25 years). The next step is a technology called VDSL2 which Telstra would deploy as part of its FTTN solution and should take us to speeds up to about 50Mbps.

VDSL2 however is still not the end of the road because a new technology called Dynamic Spectral Management (DSM) is already taking shape in laboratories and will double the speeds over the copper yet again. The catch is that the prospect of DSM is effectively killed by sub-loop unbundling. Sub-loop unbundling thus spells the end of the road for improved speeds on the copper network.

The copper network is a huge asset and to waste it in such a fashion would be criminal (probably literally because Telstra shareholders have paid for that asset and are entitled to see it used for best effect).