G.fast Could Use a Boost
G.fast is under pressure to deliver. The copper-fortifying broadband technology, which many had expected to see in commercial deployment this year, is still at trial stage in several European markets as 2017 draws closer. Forthcoming "second-generation" chips and standards have been heralded as G.fast game-changers, but few European operators have announced big commitments to G.fast in the meantime. And in a troubling development for the G.fast community, European regulators are now said to harbor serious reservations about G.fast's ability to meet regional targets for high-speed broadband connectivity.
Among Europe's biggest fixed-line operators, BT Group plc (NYSE: BT; London: BTA) has quickly emerged as the main G.fast cheerleader. Using the technology, the UK operator aims to provide 300Mbit/s services to about 10 million UK premises by 2020. Having carried out G.fast trials this year, BT is now set to cover about 140,000 homes and businesses with the technology by March 2017. Others will be watching closely to see if it measures up. (See Long-Range, High-Speed Gfast Is Coming – BT)
The ambitious scheme would not have been possible had G.fast not already taken giant strides. The technology essentially works by extending the frequency range over which broadband signals travel. When it first appeared several years ago, it was designed to increase speeds over very short copper loops between homes and nearby distribution points, losing potency over longer distances. That meant rolling fiber out to those distribution points and equipping them with the necessary electronics. For an operator wanting a nationwide solution, it looked prohibitively expensive. (See G.fast: The Dawn of Gigabit Copper?.)
on October 18-20 in London.
Thanks to progress since then, BT is now confident that G.fast can be effective over much longer distances than was originally thought possible. Instead of introducing the technology at distribution points, BT plans to install it at street cabinets, usually around 300-350 meters from customer premises. Because BT maintains as many as 4 million distribution points, but just 90,000 street cabinets, the economics of a cabinet-based deployment clearly make sense.
The G.fast workout
To realize its vision, however, BT needs G.fast to hit the technology treadmill and become even fitter. Encouragingly, the G.fast industry is currently in the midst of epic workout that will "vastly change what G.fast looks like in the next 12 months," according to Eric Joyce, a systems engineer at broadband vendor ADTRAN Inc. (Nasdaq: ADTN), which has been heavily involved in BT's G.fast trials.
Even before an "Amendment 2" became fully standardized this year, the industry had started experimenting with a reduction in what is known as the SNR (for signal to noise ratio) margin to boost performance. With G.fast, the original recommendation was for an SNR margin of 6dB to provide a sufficient "buffer" against unwanted noise, as Joyce describes it. That margin has turned out to be surprisingly robust, though. By reducing the SNR margin to 3dB, and removing some of that buffer, Adtran has seen bandwidth improvements of between 50 Mbit/s and 100 Mbit/s, depending on loop length, with "negligible impact on stability."
Next page: Second generation