Within the debate of the Multi Technology Mix of the Australian NBN, where the HFC also gets included in this mix for approx 20-25% of the connections, it interesting to follow the debate on the upgrade of this infrastructure in America.
Cable networks around the world are now faced with a major technology upgrade known as DOCSIS 3.1 and also in Australia the NBN company has indicated it is also in their plans to upgrade the 20 year old networks they have bought from Telstra and Optus along those lines.
However, as more and more information becomes available and the first companies involved in such upgrade are getting some first hand experience of the work that is needed and they see the first results of the upgrade, an interesting discussion is taking place regarding the costs involved in relation to the upgrade of the HFC networks in comparison to the cost of building FttH networks. With the new results available to them, the argument now is that in many cases the cost of a full transiting from DOCSIS 3.0 to 3.1 are so great that it would be more cost effective to go to FTTH.
The discussion around this develops as follows. Initially it was thought that upgrading the HFC network simply involved the bonding of channels in the downstream that are being freed up in the new standard by going to MPEG4 or that it simply would be a matter of turning off analogue part of the network.
However, it has become clear that just doing that is rather pointless. Because this is not going to make the service much faster than it already is, and at some point, there is not enough upstream capacity to enable the operator to actually use all that downstream bandwidth.
DOCSIS 3.1 does support upstream capacity with much more bandwidth, but to do that, operators have to increase the amount of spectrum allocated to the upstream service, so it is more symmetrical. Right now, in North America, the HFC cable operators run ‘subsplit’ upstream: 5-42 MHz on the upstream, and 50-860 or 1 GHz in the downstream. Changing that split requires changing every diplexer in every active component in the network.
This is a very substantial amount of work. If operators take the upstream beyond 75 MHz, then the set top boxes (STB) will also have to be changed. And this all assumes that TV’s are no longer plugged directly into the cable, otherwise they will need a low pass filter installed, because the cable modem transmits at up to 1000x the power of the downstream, and if they go past 50 Mhz the hi-pass filter on the TV won’t protect its frontend against a signal that strong.
Not only does this upgrade create a large amount of work it will also potentially causes a large amount of network downtime as the HFC plant gets rebuilt.
So the logic is that if operators are going to try and fix the upstream element of the network (this is the Achilles heel of the HFC network), it could add up to be more expensive than an FTTH overbuild, particularly if it is possible to overlash the fibre on the existing HFC plant and if it are aerial drops to the home.
In the underground case, it is less compelling unless you can somehow replace the copper with fibre in an easy way. Most of that wasn’t pulled through conduit, so in that case it is a nasty and expensive problem. In the US, multi dwelling unites (MDU) also don’t have anything more than coax for inside wiring (the copper pair is too slow), so going FTTH there means a lot more cost to rewire the building. In many cities, these last two issues (MDU inside plant and underground cabling) probably means the cable operator will have to spend a lot of money and as a result we see that the operator while adding FttH to its service will still opt to keep the HFC in place rather than go to a full FTTH service. There are indeed several examples of this in the USA.
So in the short term at least, most cable operators will do the “light touch” approach, because they have not signalled to the market a capex profile that can afford anything but that. And they hope that this will be good enough to fend off gigabit networks being deployed by competitors.
If that works, it could be enough to keep sufficient number of subscribers on their HFC network and they can at least for now avoid the FttH overbuild. According to some, this is the best case scenario for them – and in cases of little or no competition they even could avoid deploying the DOCSIS 3.1 light touch upgrade all together.
If however, these tactics are not going to work, and they find themselves losing share to FTTH competitors, then things get more difficult for the cable operators. They will than have to make decisions if they can take the subscriber losses and still be in business, or they need to do what ultimately is unavoidable and upgrade to FttH.. Being aware of what needs to be the final decision, some operators are now doing part overbuilds, in cases where there are enough people prepared to pay a premium price for a FttH service based on a minimum 2-year contract basis.