The Herald Sun’s chief technology writer, Andrew Bolt, has had another go at breathing life into the “wireless will do everything” argument against the National Broadband Network.
4G may trump Gillard’s $43 billion
MIT has news that cast even more doubt about the wisdom of the Gillard Government committing $43 billion to a national broadband network that leaves users chained to a socket in the wall:
I hate to repeat myself, but it seems that this fanciful idea simply will not go away by itself, so let’s take another look at it it for the benefit of Andrew and anyone else who is placing their faith in the magic wireless that’s going to avoid the laws of physics.
Let’s begin with the article that Andrew linked to. The first thing that leaps out at you is the speeds that are being discussed:
Bob Azzi, senior vice president for network at Sprint Nextel, claims that his network can offer speeds from three to five megabits per second for downloads—and one or more megabits per second for uploads.
Here’s some interesting info for Bob and Andrew, mobile broadband at that speed, and faster, is already available in Australia, yet very few people are opting to use it in preference to fixed line broadband as their primary internet connection. Why are people ignoring the utopian wireless future? Well for starters there’s the fact that wireless currently costs more than ADSL or cable, while, in Telstra’s case, delivering less than a quarter of the amount of data.
Another problem with wireless, even the much vaunted future standard LTE, is that it will still be delivering asynchronous data speeds. You can’t upload as fast as you can download, which is problematic for applications like video conferencing or for delivering large files from the user.
Then there’s the fact that wireless performance degrades every time an additional user comes online, and can also be adversely affected by the weather. As Stilgherrian pointed out in Crikey a few months back, to overcome the contention problems faced by wireless you’d need a mobile phone tower on every street corner, each of which would, incidentally, require an optic fibre cable going to them to provide them with their link to the internet.
It’s also worth addressing the idea that the NBN internet connections will leave us all “chained to a socket in the wall” as Andrew fears. From Andrew’s linked article again:
Stores typically need fast networks at checkout counters to look up product information, reference customer data, and process credit cards. “Structuring a store is much easier when you are not constrained by having to get wires to every last place,” he adds. “You can use a wireless 4G modem instead.”
That’s true. However, what is also true is that local wireless solutions have existed for some time and are probably better than a 4G modem in almost every case because the data never needs to leave your corporate network. Broadband modems with wireless capability are a consumer focussed item that have been embraced by plenty of households and businesses, and the NBN won’t change that.
There’s no argument that mobile data usage is increasing at a rapid rate, nor that it will continue to improve speed and reliability, but radio waves that have a constrained frequency spectrum will always be slower than optic fibre connections. There is no practical way that wireless can provide the large scale service that is being planned with the NBN at anywhere near the same performance. Those pretending otherwise all seem to have an investment in wireless or a political axe to grind about the NBN, neither of these make compelling technical arguments.