ADTRAN Warns G.fast Ultrafast Broadband at Risk of Perception Problem

Historically hybrid-fibre broadband services like VDSL (FTTC) rely on a mix of fibre optic and twisted-pair copper cables, although signal attenuation over the latter always ends up reducing performance over distance and the new G.fast (ITU G.9700/9701) technology is particularly susceptible.

Openreach (BT) currently intends to roll-out G.fast technology to 10 million homes and businesses across the United Kingdom by 2020. The top speed of the new service is 330Mbps (50Mbps upload), but you’d have to live within 350 metres or less from your local PCP street cabinet (G.fast pod) in order to stand a chance of getting that.

After that the speeds fall away rapidly to around the 100Mbps mark at 500 metres (note: a slower 160Mbps product will also be offered).

However ADTRAN, which supplies some of the kit being used in G.fast deployments around the world, is concerned that the benefits of the new service could be weakened if infrastructure providers and ISPs fail to make a clear product distinction between performance on long and short copper lines.

Ronan Kelly, ADTRAN’s CTO, said (Total Telecom):

“Unfortunately there’s no differentiation from the industry perspective of long-loop G.fast versus short-loop G.fast, it’s all just called G.fast. So if there’s a bad experience on one extreme, it tends to tarnish every use case.

G.fast as a technology has a Hell of a lot more to give, we’ve got to be careful of the use cases so that we don’t damage its reputation. Combining the two use cases together … I think is putting the technology at risk”

As usual the big challenge will be in how best to segregate out the two use cases and promote that to consumers in a clear and simple way. The advertising of broadband services in the UK is already a very tedious subject (example) and most ordinary consumers are unaware of their own copper loop distance (it’s also difficult to know the precise loop length of every individual line).

The previous generation of Fibre-to-the-Cabinet (VDSL2) services somewhat dealt with this by having two primary product tiers (80Mbps and 40Mbps) and personalise speed estimates on checkout (not all ISPs do this), although that didn’t completely solve the problem and people can still much slower speeds than estimated on FTTC.

Lest we forget that plenty of other factors, such as poor home wiring, EM interference and cable damage.. among other things, can also impact the performance of such services. Suffice to say that we’re not sure if there’s a simple solution to this problem, although advocates of pure fibre would no doubt say, “forget all the copper, just do FTTP/H!” or something like that (somebody would have to foot the huge bill for that of course).

Alternatively Openreach could build lots of smaller G.fast nodes and put them much closer to homes via the Fibre-to-the-distribution-point (FTTdp) approach, which would deliver a more reliable level of performance. However that seems to have been put on ice for a bit in favour of cheaper street cabinet extension pods (we still expect to see FTTdp return in the future).