SPECIAL REPORT: Upgrading the Coalition NBN

Building an affordable yet future-proof NBN is a riddle that the Coalition is yet to solve. But there is a way that the network can be built cheaper, delivered on time and optimised for upgrade.

The National Broadband Network is a vital nation building project that has been plagued by ongoing criticism of the network design, cost blow-out claims and ongoing construction delays. Despite its grand ambitions NBN Co has managed to connect far fewer premises to the NBN than originally identified in the first NBN Co rollout plan.

The Abbott government took its NBN plan - promising a faster, cheaper rollout - to the polls and has now acted quickly to reconstitute the NBN Co board and get the critical NBN-related reviews off the ground.

In the weeks prior to and immediately after the election during visits to Cairns, Darwin and Geraldton, it became apparent that the NBN rollout is now in full swing around the nation. Being able to see what’s happening on the ground outside capital cities provided a new perspective on the NBN and how it will affect our everyday lives.

During the site visits several NBN Co personnel and subcontractors mentioned that they were being asked the same question over and over by people passing in the street: “will the NBN get to my street [or suburb] before the election?”

Australians want the NBN but how does the government deliver on its promise to deliver the NBN faster, more affordably and sooner?

Design for upgrade and flexibility

There have been suggestions made that future wireless or copper based technologies will make the need for an all optical NBN unnecessary. It’s timely to remember that universities and industry research teams around the world test technologies that typically won’t come to market for 15 to 20 years.

Unless there is a sudden and unexpected break-through in wireless or copper based technologies the only known approach to provide high speed broadband that incorporates traffic class management and Quality of Service (QoS) relies on the use of optic fibre.

A smarter approach to FTTN

So what are the possible engineering solutions available that will permit the government to achieve its pre-election plan and at the same time provide an optimal low-cost upgrade path to an all optical FTTH NBN?

To whittle down the possible engineering solutions it's necessary to introduce key design criteria for an optimal low-cost upgrade path from ADSL2 , Fibre to the Node (FTTN), Fibre to the Basement (FTTB) and Hybrid Fibre Coax (HFC) to FTTH.

A list of the key design criteria highlights two build requirements for an optimal low-cost upgrade path that’s compatible with the existing NBN and provides flexibility for telcos, service providers and customers.

To achieve the two build requirements, utilising the key design criteria, the need for an engineering solution became apparent. And there is a way of making this happen.

A new NBN design

The current fibre rollout past premises should continue unabated because the fibre really needs to go as close as possible to premises.

Equipment vendors could be invited to supply devices that plug into the fibre multi-ports that provide ADSL2 /VDSL2/VDSL2 with vectoring. The ADSL2 /VDSL2/VDSL2 with vectoring customer premises modem or an additional plug pack located in premises would provide power over the copper cable to the xDSL device located in the street.

The xDSL device may include one or more ports where each port would be powered by an active connection.

The current NBN design to utilise Gigabit-capable Passive Optical Network (GPON) and a passive fibre design configuration should be retained and this means the ADSL2 /VDSL2/VDSL2 with vectoring devices that plug into the fibre multi-ports would need to convert the ADSL2 /VDSL2/VDSL2 with vectoring traffic to GPON before it entered the fibre.

When G.Fast becomes available equipment vendors would be encouraged to add G.Fast to the devices.

An alternate solution is to transfer the xDSL/G.Fast traffic up the fibre utilising a separate light wavelength to that used for the GPON network, but this is less elegant.

Benefits of a flexible upgradable NBN design

The engineering solution outlined provides an upgrade path from ADSL2 to VDSL2 to VDSL2 with vectoring to G.Fast and finally to GPON. The approach would utilise copper but limit the length of the copper used to provide for optimal xDSL speeds depending on the state of the copper rather than attenuation due to the length.

What this means is that telcos, service providers and customers are provided with flexibility and a low cost approach to upgrade from one technology to the next.

For those people who do not want to wait and would like GPON immediately they can either install the fibre themselves or pay a local contractor to do it for them. The cost of utilising a contractor would be about $1000 to $1500 based on figures from a recently leaked NBN Co document made available to Business Spectator.

Service providers are likely to offer plans that incorporate the cost of running fibre into premises. This would not be an unusual move as customers often pay for included handsets on mobile phone plans where some of the handsets costs upwards of $800.

For premises with poor copper transmission speeds may fall back from G.Fast or VDSL2 to ADSL2 then ADSL2 and finally to ADSL in the worst case. However, it is better to have graceful degradation than to try to force VDSL2 with vectoring over defective copper.

Cabinets are not needed

By adopting this engineering solution there would be no need to put 80 to 100,000 large cabinets on streets around the nation that would be filled with batteries and equipment needing large amounts of power drawn from the local supply.

There would also be no need to dig up streets to redirect copper from the existing pillar points to the cabinets. Carving up the streets becomes necessary because the copper tails from the pillar points are typically up to 750 meters in length whilst VDSL2 with vectoring is only practical over about 300-400 meters of copper based on UK experience.

There would also be a need to run fibre down the street to the cabinet so if the streets are to be dug up why not continue with the GPON fibre rollout passed premises or into premises basements?

A key aspect of the design is there should be no need to re-negotiate with Telstra for the purchase or use of copper. The portion of the copper tails to be used resides primarily on premises and property owners are likely to be able to use the copper in any way they see fit if it is disconnected from the Telstra network.

Consider all options

The simple approach for industry would be to duplicate the FTTN rollout carried out in the UK. Recently the Communications Alliance, which is the telecommunications industry body that facilitates industry-based solutions, commenced a working committee to look at the introduction of VDSL2 and Vectoring for the NBN. A review of the working committee documentation and guidelines identified that the working committee is focused on a UK style FTTN solution.

That solution requires the installation of street cabinets with power, batteries and equipment for decades to come. The problem is that the UK FTTN model is not compatible with a future GPON FTTP network and this means there would be a considerable cost to upgrade the entire nation from FTTN to FTTP.

It’s important that the government give the telecommunications industry guidance and insist that the industry consider every option if FTTN is to be introduced into the NBN.

There are more than 10 million premises in Australia and this is sufficient for equipment vendors to build devices that contain ADSL2 /VDSL2/VDSL2 with vectoring/G.Fast (in the future) and convert to GPON before connecting to a multi-port on a GPON fibre network.

The government has an opportunity to redirect the NBN without stopping the existing fibre passed premises rollout and achieve goals put to voters before the last election. More importantly. it will be able to provide an upgrade path that is compatible with the existing NBN and flexibility for telcos, service providers and customers.  

Mark Gregory is a Senior Lecturer in the School of Electrical and Computer Engineering at RMIT University