G.fast’s road toward bringing fiber-like speeds to existing copper-fed DSL networks took a step forward Wednesday as the International Telecommunications Union granted “first-stage” approval to the emerging standard.
That technology is embedded in a budding standard called G.fast, which was granted “first-stage approval” Wednesday by the International Telecommunications Union, which expects to finalize G.fast “as early as April 2014.”
ITU expects to finalize G.fast “as early as April 2014,” but this week’s first-stage stamp should give chipmakers, modem vendors and network gear suppliers the technical roadway necessary to develop G.fast products and present telcos with a platform that is designed to deliver aggregate capacity of 1 Gigabit per second that can be flexibly allocated to a telco’s downstream and upstream pathways.
“The core technology is pretty much set,” said Michael Weissman, vice president of marketing for Sckipio, a G.fast silicon startup that this week announced it had secured a $10 million investment from Gemini Israel Ventures, Genesis Partners, Amiti Capital, and Aviv Ventures. Sckipio, a company with about 25 employees, expects to have some initial G.fast product available in 2014. “We have trials scheduled,” Weissman said.
G.fast, which has also attracted interest from Broadcom, will flirt with 1-Gig speeds thanks to its use of a wider swath of spectrum. The standard calls for G.fast to handle bandwidth profiles of 106 MHz and just north of 200 MHz, comparing to the relatively paltry 17 MHz of bandwidth used for the 17a profile of VDSL. Because G.fast can operate in higher frequencies, it will aim to reduce crosstalk interference through the use of vectoring, a technique that is in widespread commercial deployment on DSL networks.
But G.fast isn’t a complete slam dunk. Telcos that utilize it will still be required to pull fiber to within 250 meters of the premises under an architecture referred to as Fiber To The Distribution Point, or FTTdp. In a recent report about the budding technology, Broadbandtrends LLC analyst Teresa Mastrangelo predicted that the “majority” of G.fast loop lengths will be in the range of 30 meters to 50 meters.
Although G.fast requires that telcos pull fiber closer to the premises, the technology intends to prolong the life of copper-fed DSL plant while being less complicated and less expensive to deploy than FTTH. Sckipio, which of course has much to gain if G.fast takes off, estimates that the approach will cost telcos less than $300 per home, versus more than $1,500 for FTTH.
If it works as advertised, it could also give telcos another weapon to wield against cable’s DOCSIS platform, which is already capable of hitting 1 Gbps in the downstream with DOCSIS 3.0. The new DOCSIS 3.1 spec is aiming for multi-Gigabit speeds.
“Fiber to the home is a nice response, but it really never pays for itself. They [the telcos] need something that is affordable,” Weissman said. “Cable has the edge because VDSL has struggled to keep up with the performance requirements of DOCSIS 3.0, and the telcos need a response.”
More about G.fast developments, including details on recent trial activity involving pre-standard G.fast products, will be covered in the December 16 issue of Multichannel News.
