G.fast Coming Faster and Faster

The ability of copper to transport data at fast enough speeds to keep them in the modern telecommunications game is a great thing for the telcos, who of course are saddled with immense amounts of the stuff in their networks. The availability of clever technology that transforms a burden to an asset is the telco version of hitting the lottery.

And it is happening. The latest version of the basic technology used to push copper, digital subscriber line (DSL), is being studied in Europe for use as a way to backhaul signals from cell towers to points within the network. The new approach, called G.fast, is the subject of the Gigabits Over the Legacy Drop (GOLD) project.

Things must be going well: Backhauling cellular traffic is a vital and exacting operation. The consortium features 14 organizations from nine countries. The latest project was preceded by a lot of work:

GOLD builds on the success of the completed HFCC/G.fast project, which demonstrated throughput of nearly 1Gbps per copper pair at 100 meters, and up to 170Mbps per copper pair at 480 meters, on a 16 pair standard cable. This is as much as an order of magnitude improvement compared to existing DSL technologies. GOLD will push G.fast even further to multiple-gigabit copper access rates by exploring a second version of the G.fast standard working at higher frequencies and preparing the ground for fifth generation fixed broadband.

A very important point stands out in Mary Lennighan’s story at Total Telecom on the emergence of G.fast and the GOLD project. Fiber-to-the-home is expensive and offers high levels of bandwidth. That’s well understood. What isn’t as commonly considered is that demand can lag and these projects can be overkill:

Operators in Japan and Korea were ahead of Europe in rolling out high-speed networks, but “the stickiness of services in those countries is not what people think it is,” warned Robin Mersh, CEO of the Broadband Forum. In some cases, people have signed up to 100 Mbps broadband only to later downgrade to 50 Mbps because they don’t want to pay for what they are not using, he said.

The idea that demand over the long haul may not be as great as anticipated makes an approach that relies merely on upgrading existing infrastructure – and avoids the expense and headaches of “rip and replace” projects — even more attractive.

There are other signs that the era of G.fast is upon us. For instance, Siklu and Sckipio Technologies, millimeter wave and G.fast chip companies, respectively, announced at the Mobile World Congress earlier this month that they will provide broadband to dense urban environments. The plan is to offer data at rates as fast as 2 Gigabits per second (Gbps) for 2.5 miles or less. The system will use Siklu’s E-Band and V-Band and Sckipio’s G.fast technology, the release said.

Lightwave reports on a platform that combines Gigabit passive optical networks (GPONs) with G.fast. The Hybrid Fiber-Copper Access Distribution Point (HFCA-DP) is from Telnet Redes Inteligentes S.A.; Sckipio is providing the chipsets.

Both of the introductions take the step of replacing coaxial cable or fiber in the last mile. That suggests that vendors and service providers are getting on the G.fast ship.