In my last post on the NBN I got a few things wrong. The $41 billion pricetag is not just for the satellite service. The satellite component is priced at $2 billion (or $3.5 billion if you include fixed wireless, or anything up to around $5 billion if you actually want all households to be serviced, according to this report - page 13, available on the NBN Co website). Does this change the base argument? Not really. If you look into that report a bit further you find some interesting gems, but first, let's look at some NBN marketing. In this piece we are told just how wonderful the NBN is, but there are a few asterisks in that text that then point to this:
*Your experience including the speeds actually achieved over the nbn™ network depends on the technology over which services are delivered to your premises and some factors outside our control like your equipment quality, software, broadband plans and how your service provider designs its network.
So back to the report. It states that even at 65% take up of the satellite service, the "mean busy hour throughput" will be 150kbps - or just about 3 times dialup speed. Hang on, isn't it supposed to be 25Mbps down, 5Mbps up? Well, here is where the marketing and the technical realities diverge. Unlike FTTP where you have a high-bandwidth connection that comes directly to the house, with satellite you have a limited number of transponders, each of which supplies a beam that covers a fixed geographical area. Each transponder is limited in its bandwidth, and that bandwidth is shared amongst the 500 to 15,000 premises covered by that beam. So if we actually use the service, the speed we get is determined by where we live and hence how many other premises are covered by the same beam. The more successful the marketing, the less useable is the outcome.
The interim satellite service (ISS) quickly became over subscribed as mentioned in the previous post, and the above report suggests that this is likely with the new service. Indeed one of the options it recommends is a third satellite coupled with expansion of the fixed wireless network. I live in central Victoria, and if you take a look at Exhibit 14-2 on page 87 of the report you will see that most of Victoria at just 65% uptake is likely to be oversubscribed.
One other fact coming from this report is that the service is not built to be future proof, and the timelines for getting people connected are starting to become very concerning. It is unlikely that the satellite/fixed wireless rollout will be completed before 2021, and even then the rate of growth of the customer base is likely to outstrip the supply by between 10 and 30% depending on the speed options being supplied (Exhibit 15-1 page 98). The satellites themselves have a design life of 15 years, so we need to revisit this before then at any rate. Yet the obvious choice of upgrading exchanges to support ADSL as an interim is simply not even considered.
Time for a bit of a reality check. Every town that has a school in it has an ADSL enabled exchange. Yet you cannot get ADSL in many of those towns. In the town just 10 minutes from here they have ADSL at the school, but no other premises has been able to get this with the excuse that there is no ADSL capacity at the exchange. Seriously? ADSL units come in various configurations, but usually start at 12 ports and go up from there. That means that there should be at least 11 other premises capable of being connected to ADSL, yet none can be.
Let's go back a step, in the Statement of Expectations signed by the Malcolm Turnbull - Minister for Communications and Mathias Cormann - Minister for Finance, the expectation is set that:
The Australian Government is committed to ... ensuring all Australians have access to very fast broadband as soon as possible, at affordable prices, and at least cost to taxpayers
There is also a recommendation in the report that NBN set expectations correctly. So why are we still getting the lie that we will have 25/5 Mbps speeds which will, in the main, be unattainable and that the NBN will be the same for all when quite clearly it cannot be? Why are obvious solutions that at least give "metro equivalent" capacity not being used? When will the digital divide really be a thing of the past?