- Widely available 500 Mbps speeds over copper remain several years away, but a handful of European incumbents are looking at the technology very carefully as a realistic means to get high-performance access economically leveraging existing copper plant.
- The inflection point for widespread availability is going to hinge on sensible market regulation, so that domestic incumbents are encouraged to invest in G.fast while alternative ISPs and telcos have incentives to deploy G.fast-based offerings via LLU.
It is amazing to see how humble copper continues to guarantee its own survival based on good reasons to maintain it in the face of growing optical fibre investments. Last July, A1 Telekom Austria and Alcatel-Lucent achieved 1.1 Gbps speeds over good quality cable along a distance of 70 metres and 800 Mbps over 100 metres in test cases. The ITU-T founded the G.fast group in 2011, and the program has the support of several operators, chipset manufacturers and equipment vendors. But, here’s the rub: Rollout is bound to become a contentious issue, fraught with legal battles between incumbents and alternative providers and overseen by domestic market watchdogs. When it comes to deploying equipment and chipsets supporting G.fast, why would the incumbent invest only to be forced to hand the keys to the treasure chest to rivals ISPs and telcos? Let’s leave the legal battles to one side; suffice to say, this will continue to keep the legal heads in Brussels, and throughout Europe, busy for a while to come. Luckily, the early phases of DSL and the enforcement of LLU have formed a foundation, and G.fast regulation will surely be built upon similar models. Across Europe and the rest of the world, copper pair quality, distances from exchanges and construction also vary wildly, so there will only ever be a patchwork quilt rollout for G.fast, with differences a certainty country by country. Wholesale divisions, such as BT Wholesale, may play a fundamental role in developing white label versions that alternative telcos and ISPs can then take to a wider national market, which will help to encourage a healthy competitive landscape that ultimately benefits the buyer.
TeliaSonera says it has deployed G.fast to a set of end users in Helsinki, based on Huawei equipment. There are not many details available on the deployment, such as the number of users; so far, this is all pretty hush-hush. Typically, operators are relying on FTTC investments to get higher speeds into populated neighbours and business areas, but the last mile piece will still have to rely on copper pair. The potential of G.fast means that smaller businesses outside of the fibre investment focus and without the funds to source 100 Mbps fibre or leased lines have a good chance of getting higher-speed accesses over budget-priced copper in around four to five years’ time. However, this will continue to depend on geographical location, so some rural areas may continue to struggle due to distances from exchanges. (Sorry, but there is no simple and magical fix there.) For the IT manager, G.fast – when it comes – means more capacity to run corporate video and share images, better performance for all things running in the cloud, including storage and back-up, and the ability to make flexible changes to computing resources on the fly. In the meantime, VDSL, DOCSIS 3.0 and Ethernet access services will continue to provide the backbone pillars for mid-band access speeds supporting the enterprise, and some patience is called for until higher speeds over fibre and other mediums come to market.
BT’s fibre-on-demand service is also an interesting option. It is in the sub-GBP 1,500 annual subscription range and provides speeds of 330 Mbps downstream and 30 Mbps upstream. The costs for installation are up to GBP 6,000 in most cases, meaning a one or two-year return versus the costs of leased line/Ethernet, which represents a good return versus EFM in the medium term. The final point which may stifle interest in G.fast is that the role and availability of 4G will grow massively also over the next few years, but we have seen that fixed and wireless accesses tend to cooperate and complement as opposed to cannibalize each other.