The political message of the Coalition’s Plan for Real Action on Broadband and Telecommunications is that they can deliver the same outcome as Labor’s National Broadband Network (NBN) for $6 billion instead of $43 billion, sooner and with less risk. But while both Labor and the Coalition promise headline speeds of between 12 and 100 megabits per second (Mb/s) for 97% of the population and satellite for the rest, they’re otherwise wildly different propositions.

A broadband network is, simplistically, these pieces. A customer access network (CAN) connecting everyone to the local exchange. A backhaul network running between cities connecting the exchanges together. A core network that ties it all together, connects to the rest of the world, and provides administrative functions such as metering. Together these constitute the wholesale network.

On top of that you need retail operations such as marketing, sales, billing and customer support.

Before we go further, if you need to brush up on “megabits” and what’s possible at different speeds, try our NBN Clarifier part 1.

Now for a century or more the CAN has been copper wire. Mostly that’s still the case. Currently the fastest data over copper is ADSL at up to 24Mb/s. However speed declines rapidly further from the exchange, and the maximum distance is just a few kilometres. Physics is physics, and this is pretty much it.

In some areas line-sharing techniques such as pair gain and RIM mean that ADSL is limited to around 1Mb/s or can’t be used at all.

In parts of some cities there’s also hybrid-fibre coax (HFC), the “cable” networks laid down by Telstra/Foxtel and Optus, originally for pay TV. This has optical fibre cable from the exchange to a neighbourhood node that typically serves a few hundred homes, then shielded coaxial cable. Maximum speed is up to 100Mb/s using the DOCSIS 3.0 standard that both Telstra and Optus have been rolling out, with the potential for a few multiples of that, but around 10Mb/s otherwise.

Then there are fibre CANs.

Fibre to the Node (FTTN), as in Labor’s plan at the 2007 election, uses cable or DSL for the last segment. End-user speeds are therefore broadly similar to HFC, or to ADSL when you’re close to the exchange.

Fibre to the Premises (FTTP) is fibre all the way. It typically provides 100Mb/s now, but can be upgraded by changing the devices at each end. Indeed, technology already exists to run fibre many orders of magnitude faster, although at higher cost. For example the forthcoming Pacific Fibre/Pacnet cable from Sydney to Los Angeles via Auckland will run at around 2,500,000Mb/s on each fibre pair.

There are also rapidly-evolving wireless technologies. Without getting into the messy details, the fastest are currently Telstra’s Next G, a 3.5G network offering up to 24Mb/s, and the vividwireless 4G network which can provide up to 100Mb/s to mobile devices and 1000Mb/s – that is, 1 gigabit per second (1Gb/s) – to fixed locations. Faster will come, sure. But wireless has physical limitations, detailed in Crikey’s second NBN Clarifier, which can’t be ignored.

Throughout all this sits that magic phrase “up to”. That’s where the parties’ policies differ.

A fibre is yours all the way back to the exchange. With HFC you share the bandwidth back to the node. With wireless, you share back to the tower.

FTTN also has the potential to run symmetrically — that is, upload speeds as fast as download. This means you can, say, send high-definition video as well as receive it. ADSL2+, HFC and wireless all run asymmetrically, with upload speeds significantly slower than download. Fine for watching content produced by someone else. Not so good if you wan to create or participate as an equal.

Labor’s NBN is about completely replacing the copper CAN with FTTN for 93% of the population, with 12Mb/s fixed wireless for another 4%. They’ve published maps showing where the boundaries lie. FTTN is an expensive infrastructure build, but higher-profit areas like the central cities cross-subsidises the regions. It provides a consistent network nationally that can continue to be upgraded for decades.

The Coalition’s policy is to primarily leave things to the market, but focus on areas where the market has failed. The Fixed Broadband Optimisation program would upgrading sections of the copper network to remove pair-gain, RIM and poor-quality wiring so ADSL2+ can be provided, and speeding up HFC where it exists. Carriers would be subsidized to build new rural and remote wireless networks, starting with $200 million in 2011-12. New metropolitan wireless networks start with $375 million in 2013-14.

The Coalition has so far avoided saying who’ll get which technology where, and who gets 12Mb/s rather than 100Mb/s. So far they haven’t said what spectrum will be used for the new wireless networks either.

When it comes to backhaul, the policies are similar. In areas where there’s only one backhaul provider, usually Telstra, competing backhaul will be constructed. But the NBN is already doing it, whereas the Coalition starts with a $50 million spend in 2012-13.

But choosing one plan over the other is as much about ideology, vision and political rhetoric as technological choice. Labor is spending a lot now to build for the future. The Coalition is spending less on its “affordable broadband” — which certainly saves money now, but is it merely delaying the inevitable big spend?