Fibre to the Node (FTTN)


FTTN Layout

FTTN Layout

An FTTN system consists of powered cabinates connect by fibre to existing exchanges. Premises are connected to the cabinets using the existing copper network.

In a FTTP implementation fiber runs all the way from the central office to each home and is clearly the end objective.  FTTP architectures are future-safe, enabling operators to deliver (more or less) unlimited bandwidth: 100 Mbps per household today, with technology evolutions set to increase this in the future. Their main drawback, however, is that (nationwide) FTTP deployments require a lot of time and money.

This is exactly where an FTTN solution  can look attractive.  FTTN can be installed quicker and will initially cost less. It can later be converted to a full FTTP solution.

Typically, in existing buit-up areas (“brownfields”), initial  FTTN deployment will much cheaper than an FTTP architecture because the copper between the street cabinet and the subscriber is being re-used.  This is also the reason why it will be faster to install. To build new suburbs (“greenfields”) with FTTP is undeniably the best solution and this is wat both Australian political parties propose.

For brownfields FTTN permits a less capital-intensive deployment than FTTP.  It allows leverage to the largest extent possible of the existing copper, and enable operators to start offering premium services quickly. It  can start to generate revenue, all while getting more time to roll out their fiber network.

In the Australian environment, the benefits of FTTN are outweighted by the disadvantages. FTTN nodes have to powered.  Depending on how many premises will be served, it may require up to 70,000 powered nodes. The cost of these is not insignificant. If these nodes will require an upgrade to FTTP in the near future to meet expected bandwidth demand, this will be money wasted.

FTTP nodes do not require power and will also require less maintenance.  When the FTTN nodes are converted to FTTP, the power is not required.  Much of the FTTN equipment will be redundent.

The quality of the existing copper lines from the nodes to the premises is  major factor which influences the speeds obtainable with a FTTN configuration. Much of the existing copper has been in the ground for more than 50 years. Unless these copper lines are replaced, a FTTP solution will not be able to provide the theoretical speeds of 100Mb/s for such systems and the communication of these copper lines will be no more relaible than the current network. In any case, the speeds obtainable still reduces for premises that are long distances from the nodes. To replace the copper would be a futile exersize, given that its replacemant with fibre in the short term will be required in any case. Below is an extract from a article that explains how distance from the node effects speed in a FTTN solution:

The best-documented FTTN plan in Australia remains Telstra’s November 2005 strategy document, described in this presentation. That proposed installing 20,000 nodes within 1.5 KM of 4 million addresses – for a mediocre outcome of just 12 Mb/s guaranteed, capital city coverage only, over three years, and at a cost of between $3 billion and $4 billion.

To get the loop down from 1.5 KM to 600 meters (achieving 50 Mb/s if the copper is in good condition) or 300 meters (for 100 Mb/s performance) needs many, many more nodes – as many as 50,000, perhaps as high as 70,000. Each of these nodes needs electricity (today, the copper network mostly supplies a mere trickle of electricity to standard telephones – not Kilowatts to power big kerbside boxes) for power and cooling.


Replacing the current CAN and RIM networks with a FTTN system requires that placement of the FTTN Nodes should be in the same position where the CAN pillars and RIM cabinets currently are. If not, the existing copper will also have to be replaced.  This is not necessarily the best position for a FTTP network. A new FTTP system  have more flexibility in positioning the nodes because all the copper has to be replaced anyway. Nodes can be placed in positions which take into account the requirements of FTTP networks, as well as changes in which occurred in brownfields over the years, eg new infill areas and old homes having been replaced with newer and often higher density buildings.  Having to re-use the existing ducts to a large extend to save money does reduce the flexibility for configuring FTTP systems as well.

With the existing CAN network Telstra was forced to provide access to the network by letting ISPs place equipment in Exchanges that connects to the copper lines. This “unconditioned local loop unbundling” (ULL unbundeling)  or “facilities based unbundling” provided some  competition but there was constant friction and complaints from ISPs who felt that they did not get a fair deal from Telstra.  The NBN solves this problem neatly by providing a so called “level 2 bitstream” inteface at the 121 points of interchange. This provides a level playing field for all service providers,  including Telstra, and opens the way for many more service providers to provide new services.

It is not clear at the moment how a FTTN solution would be able to provide this unbundling.  If Telstra is again given control of the “last mile”,  all the work done to achieve the separation of the wholesale and retail portions from the network would be undone. Below is  another extract from the same article referenced above:

If the ISPs have to disconnect and discard their existing ADSL2+ equipment, the question arises: who will operate the DSLAMs at the nodes?

The only options are to put the nodes entirely in the hands of a network operator; or to allow multiple equipment owners at the nodes. The first option clearly eliminates facilities-based competition – the retailer simply buys a bitstream from the network operator, like the NBN but without the fibre.

And the second option – multiple owners competing at the nodes – is only marginally feasible, from a technical point of view. In hotly-contested areas, there would be at least four competitors – Telstra, Optus, iiNet, TPG and iPrimus – and that’s ignoring the remaining business-specialist or regional-specialist DSL owners.

With that many DSLAMs in a node, interference and incompatibility between owners is a near-certainty, unless the network operator mandated a single-vendor build.