A new report from Ovum, which was jointly commissioned by BT and Australia’s nbn, has forecast that some 29 million subscribers will be using a G.fast based ultrafast broadband connection by 2021 (3% of the global fixed broadband market) and most of those will be in Western Europe.
As an emerging technology the growth in annual subscriber additions is expected to accelerate in each year, rising from 330,000 in 2017 to nearly 11.5m in 2021.
Advertisement
The ‘Gigabit Networks‘ report reads a bit like a long promotional advert for the technology and sadly doesn’t offer much in the way of new information, although it does provide a short but useful summary of how different operators from around the world plan to implement G.fast (ITU G.9700/9701).
We also get an idea of where the uptake will be strongest and unsurprisingly G.fast will be most popular in Europe, not least because a lot of telecoms operators in Western Europe still have a large amount of legacy copper based infrastructure in the ground. This makes G.fast a much faster and cheaper way of delivering “ultrafast” broadband services to premises than a pure fibre optic network (FTTP/H).
BT (Openreach) will of course be one of the technology’s earliest adopters and they hope to begin the commercial roll-out next year, which could reach 10 million UK premises by 2020, with “most of the UK” then likely to be covered by 2025 (we’re guesstimate that “most” might equate to around 60% UK coverage). See here for the technical details.
In many ways that roll-out has already started and Openreach intend to grow their G.fast pilot to 17 UK locations from January 2017, reaching around 138,000 UK premises by the end of March 2017 (here). Initially the service will offer top download speeds of ‘up to’ 300Mbps (50Mbps upload), with a lower tier of 160Mbps (30Mbps upload).
Advertisement
In the future BT has hinted that they could squeeze a top speed of 500Mbps out of the technology, although the results from their recent trial (here) and the plan to deploy most of G.fast from an extension to existing PCP street cabinets suggests that this may be difficult to deliver; not everybody will be able to get the top speed (similar to the coverage impact of 40Mbps vs 80Mbps on FTTC / VDSL2).
G.fast requires much higher frequencies to achieve higher data rates (106MHz+), but this creates strong signal attenuation and G.fast can’t be sustained over long copper distances. The report confirms a maximum ideal of about 250 meters and significantly lower than this for the highest data rates. BT seems to be targeting an ideal distance of up to around 300-350 metres in their trials, but others are trying to keep the copper run even shorter in order to push the best speeds.
Ovum’s report notes that quite a good number of other operators have opted to take the more expensive and fibre-rich approach of using a Fibre-to-the-distribution-point (FTTdp) style model, which runs the fibre optic cable to smaller nodes on top of telegraph poles, inside buildings or under manholes etc.
For example, Taiwan’s Chunghwa Telecom have utilized a single-port unit with copper line lengths kept under 200 metres, and so enabled speed tiers of up to 500/250Mbps to be supported. Similarly Swisscom in Switzerland are also trying to keep G.fast to under 200 metres of copper from the end-user, thus allowing speeds to be lifted up to 500Mbps.
Advertisement
The report also touches on the future XG-FAST technology, which has already been tested by BT in 2015 and was able to deliver aggregate speeds of 5.6Gbps over just 35 metres of copper cable and 1.8Gbps at 100 metres (here). However we don’t get any new information on this front and XG-FAST is not yet an official standard. BT would also need a much more fibre-rich deployment in order to make the best mass-market use of the technology.
NOTE: The article picture is of an underground G.fast node (coloured white) in an FTTdp setup.
Download the Gigabit Networks Report
http://www.nbnco.com.au/Gigabit Networks.pdf
Comments are closed