FTTDU - Fiber to the Down Under

While we in Australia await the deliberations of the government's expert panel on fibre-to-the-node (FTTN), other countries in the region are steadily building fibre-to-the-home (FTTH) access networks. Ovum's estimates are that by the end of 2009 there will be approximately 13.8 million FTTH subscribers globally, 82% of whom will reside in the Asia-Pacific region. FTTH activity in the region is concentrated in Japan, the Republic of Korea and Taiwan. In the US, Verizon is undertaking a high-profile campaign of installing FTTH for its fibre-optic service (FiOS) customers.

Is FTTH a possibility for Australia, specifically for the great Australian suburbs? Currently available public information suggests a cost of around $1,500 per household. This is probably more than twice the cost of an FTTN alternative, except where there are extensive civil works (as in new housing estates). Although the costs can be lower - in Taiwan, Chunghwa Telecom is planning to spend a little less than $1,000 per premise for fibre to about 1.8 million premises over the next five years - this is still a very large investment that requires a solid business case or clear policy objectives.

There are four main ways in which a major network upgrade can be justified: investment in national infrastructure; new services; operations cost savings; and competitive response.

The first - investment in national infrastructure - has recently been revived by the Australian Opposition and has received some attention. The ability to provide higher-speed broadband services should be a boost to national productivity. In the early years, there may be little to choose between FTTN and FTTH in terms of productivity: both could deliver 50Mbit/s downstream and 1-2Mbit/s upstream. But, eventually, if we all want the line rates currently delivered to office desktops - namely, 100-1,000 Mbit/s - then FTTH will be required. If government money is to be applied to the project, then the architecture should be future proof or easily upgradeable.

The second justification - new services - is usually about delivery of IPTV, the next generation of television with many new features. IPTV is a secondary issue in Australia, where cable TV is still immature. Higher-speed Internet access, though, is a potential generator of new revenues. Small-business producers of large volumes of digital content and online financial traders are among the users who would benefit from higher access speeds and who would be likely to pay more. However, one should not expect that incremental service revenues would be sufficient to justify FTTH over FTTN - the cost difference is just too great.

The third possible justification - operations costs savings - is a key consideration for some companies, such as Verizon. An influential report in the US from 2004 suggested that the US telcos could justify an aggressive rollout of FTTH based on operations costs savings and new service revenues. This has been taken up by FTTH advocates. It is certainly true that the network-related operations costs for FTTH are substantially less than for current copper networks and for FTTN networks. There is still little experience of optical terminations in the home and it is likely, for the first few years at least, that the maintenance of optical network units in homes will be a major cost. In short, the operations costs savings are not yet certain.

The fourth possibility - competitive response - is a major driver for the telcos in the US, because of competition from the cable TV operators. This is also an issue in Hong Kong and Japan, but not in Australia. Facilities-based competition in Australia is patchy at best, although some see Telstra's proposed FTTN programme as a pre-emptive strike against DSLAMs by alternative service providers in Telstra exchanges. Any significant rollout of FTTN or FTTH will require regulatory oversight - and perhaps intervention - to provide some form of equal access for alternative service providers. FTTH raises some major issues of how alternative service providers could differentiate their offerings on an incumbent's infrastructure. FTTN, on the other hand, permits sub-loop unbundling within its access possibilities.

It looks like many Australian suburbs will get FTTN over the next few years, but only a few newly built ones will get FTTH. Does this matter? Well, it certainly entrenches relatively high operations costs for the time being - perhaps for a generation, since major access network upgrades happen rarely. The additional costs will have to be paid for through retail prices. It may also eventually limit productivity growth if the predictions for bandwidth requirements of 100 Mbit/s per household or higher come true. An upgrade to FTTH in the next generation seems highly likely.

The government response would do well, then, to promote FTTH where it is cost justified, while supporting or accelerating the rollout of FTTN. A broader view of the issues, such as reduced road traffic through greater home working and remote delivery of health and education services, could start to close the cost gap between FTTN and FTTH. Could it close the gap completely? A rough estimate based on publicly available figures suggests the government would need to provide about $4.5 billion over 4-5 years for the first 4.5 million homes to have FTTH and perhaps another $6 billion later for the remaining 3 million homes. This is comparable to support for land transport in the federal budget (about $4 billion per annum) but the political will to do it is missing.

Leith Campbell is a Principal Consultant with Ovum Consulting in Melbourne. He was previously the CEO of the Australian Telecommunications Cooperative Research Centre.