Attendees at the recent Fixed Access Networks Summit in Berlin contributed to a lively conference. In focus was the upcoming and innovative ITU-T G.fast standard and its potential, in unison with vectoring techniques, to inject new life into existing telephone wire infrastructure in the predominantly copper ‘last mile’ (between the exchange, where fibre ‘drops’, and the customer premises).
Vectoring is a means to eliminate interference, known as ‘cross-talk’, between multiple wire pairs in a single copper cable (not dissimilar to the noise-cancellation technology used in headphones). It was standardized in 2010 as ITU-T G.993.5 and the impressive results seen with its deployment in the field sparked huge upward revisions in analysts’ forecasts of the life left in copper.
Belgacom, for example, in its vectoring field trials with over 3,000 customers, reported a nearly four-fold increase in customers’ access speeds over copper connections (VDSL2 vectored, compared with un-vectored VDSL2). This suggests that copper can be seen as a realistic alternative to end-to-end fibre and cable offerings, especially in so-called ‘brownfields’ deployment scenarios such as developed cities with an abundance of copper infrastructure.
Vectoring is only part of the story though, and the advent of G.fast, in Europe to start with, looks increasingly certain. At this early stage, it is unusual to see quite so much agreement on the advantages of an upcoming standard as well as such advanced deployment plans from operators. This does suggest that G.fast is hitting the sweet spot for many.
For operators with copper assets in the access network, there are two primary reasons for G.fast within FTTdp (the Broadband Forum’s ‘fibre to the distribution point’ architecture, which extends fibre to distribution points very close to the customer premises). Firstly, it allows them to get to market more quickly with viable offerings able to compete, in ‘value-for-money’ terms, with end-to-end fibre (FTTH, fibre to the home) and the next update of the DOCSIS-based connections that offer high-speed Internet access over cable (coaxial or hybrid fibre/coaxial cable), a medium originally designed to deliver TV and sound programs to a mass audience.
Secondly, speed to market is coupled with a lower cost of deployment, making use of existing telephone wiring. There was almost universal agreement in Berlin that the cost of extending fibre to an individual customer premises can be prohibitive in variety of scenarios, such as in an apartment block with thick walls already wired with legacy copper, where FTTH deployment would be slow and expensive.
G.fast is designed for use in a ‘last-mile’ of less than 250 metres. This allows fibre to get as far as the basement of an apartment block, for example, eliminating the need to rewire the whole building and still allowing a noteworthy uplift in access speeds. G.fast requires a short loop (less than 250 metres) and operates at higher frequencies than (x)DSL, increasing the risk of cross-talk and thus making it important to deploy G.fast in combination with vectoring.
The industry is testing G.fast technology based on drafts of the standard, and the results look impressive. G.fast will coexist with FTTH, access-over-cable networks and other access-over-copper (x)DSL systems, giving rise to a wide customer base keen to have a tool like G.fast at their disposal.
All of this means that FTTH, while still the preferred choice in a ‘greenfields’ deployment scenario, has been pushed back on the timeline, for some operators at least. G.fast will help cover bandwidth requirements for some time particularly with the symmetry that the standard promises.
However, as BT has shown, it is possible to make end-to-end fibre more accessible for operators and, by extension, their customers. FTTH can be achieved, BT say, by using overhead fibre (i.e. not deployed underground) and simply getting more efficient and practiced at the job. This strategy is said to be suited to application in rural areas too, where overhead fibre deployment can prove easier, away from urban clutter.
Such developments show that rural overhead fibre deployment is now well within the cost-competitive range. It could even lead to something of a reversal of traditional network deployment patterns, with low-density areas actually getting coverage earlier than denser urban areas.
And, of course, G.fast is not yet a done deal. While progressing rapidly, it is still working its way through the approval process at the ITU. And it only takes one chewy technical issue to hinder progress and delay final sign-off.
Powering the equipment, previously a contentious issue, has been resolved. There is now general agreement that power could and would be provided from the customer premises using the remaining copper as a conduit. The only challenges left here appear to be guaranteeing enough power when only a single user is in the circuit and convincing consumers that the additional costs are worth it.
Another issue is G.fast’s requiring the unbundling of the sub-loop. G.fast’s short range and high frequencies mean that cross-talk will become an issue between copper wires in a bundle. Technically speaking, the G.fast development effort has defeated attempts to allow different ISPs (Internet Service Providers) to ‘pick-and-choose’ individual lines and clients, and to date there are no proposals that seem likely to enable this.
Different markets in Europe are adopting different approaches. Italy has allowed the requirement for unbundling to be dropped, at least for the moment, as part of its efforts to stimulate deployment. Slovenian operators are lobbying for the same treatment.
Germany has adopted another approach. Cabinets (where the fibre drops near to the customer premises) will belong to whoever activates them first. There are a number of qualifiers and special situations but a first-come, first (to) serve approach should apply overall.
Europe will continue to be a patchwork of copper, fibre and coaxial cable, certainly for the next decade and quite likely beyond. This is a healthy situation for the suppliers and the consumers. One size is never going to fit all and it makes sense to have a technology available, whether copper or fibre or coaxial, that allows choice for those selling and buying.
G.fast is the latest and greatest (in bandwidth terms), and possibly the last great leap forward for copper. Fibre will be so close to the customer premises that the next step will either be to complete the circuit with fibre, or, just perhaps, someone will come up with a way to get 40 Gbps over copper (or even 70 Gbps) out of the lab and into customers’ premises. With that sort of pipe available, we might even see copper still active towards the end of the century, not just the end of the next decade.
Send this to a friend