When it comes to broadband connections, most people will opt for cable or fiber; they’re the fastest you can get these days.
But many in the US, and other countries, are still connecting via Digital Subscriber Line, or DSL. It comes to a home, multi-dwelling unit (MDU, like an apartment building or condo complex), or even a business using old-school copper wire that used to be just for talking on the phone.
DSL is everywhere because of the phone landline infrastructure, but is hampered by the fact that the distance of a connection can slow it down. Plus, the average DSL connection download throughput usually tops out around 3 Megabits per second (Mbps) in the real world. That’s a speed that even the Federal Communications Commission doesn’t consider “broadband” anymore.
So you’d be excused for seeing a few headlines about a technology called Gfast (previously styled G.fast, as in “gee-dot-fast,” but that’s going away) and thinking that your slow-ish copper line could someday shoot your internet speeds to as high as 1 Gigabit per second (Gbps). That’s not the case, at least not entirely.
What Gfast Does
We’ve all heard of fiber-to-the-home (FttH) where companies run fiber optic cables right up to the household, but that’s not always an option.
“The biggest issue with fiber to the home [is] those last meters,” says Robin Mersh, a telecom industry veteran who’s currently the CEO of the Broadband Forum, a non-profit consortium of ISPs and equipment makers. “It’s also…the biggest drag on deployment. Sometimes you hit issues with consumers or businesses, particularly consumers, saying they don’t want a yard torn up, or holes in the walls.”
That’s where Gfast can come in. When a home or MDU already has copper lines, Gfast can take over where the DSL was. However, it has to be paired with a faster line not far from the dwelling—that’s where the fiber comes in. Thus, it becomes a fiber-to-the-node (FttN) setup. The fiber optics could come to a drop outside the building or to the basement. As long as the copper or even coax lines are in short loops of 500 meters (1,640 feet) or less, the users could get speeds between 150Mbps and 1Gbps.
“It’s not going to be a replacement for DSL,” says Mersh. But if there’s already copper in a neighborhood, it’s possible Gfast could come into play, as long as someone brings a fiber line to the area. Gfast is “a way to extend the reach of fiber technology”; because there’s less disruption at the last few meters, it’s “revolutionary” in its cost effectiveness, he adds.
Who’s Got Gfast
While the tech has been around for a few years, it’s only now getting some deployments by major ISPs.
In particular, AT&T has announced MDU deployments in several US metropolitan areas (which will include free DirecTV without a dish); currently installs are in Boston, Denver, Minneapolis, New York City, Philadelphia, Phoenix, Seattle, and Tampa, with 14 more to come.
Frontier Communications and Nokia have announced limited trials in Connecticut; CenturyLink is also giving it a go in MDUs in Wisconsin. Outside of the states, Gfast is being embraced by Australia’s National Broadband Network (nbn) through work with NetComm to make fiber-to-the-curb deployments, and BT in the UK, which promises to reach 12 million homes by 2020.
That’s pretty limited, but if the trials work out, and more fiber can be laid to the right places, it could make a huge difference for existing homes.
Where it won’t make much impact, probably, is in already underserved rural areas. Mersh still sees the issue there as the lack of fiber: “Rural remains an issue. The industry has made some strides, but some locations are particularly complex to service, and that means higher costs,” he says.
In other words, rural folks are still stuck with older, slow DSL at best, with wireless or even satellite options as backup (and the less said about them, the better).
The landscape of Gfast products is expanding rapidly, as evidenced by all the Gfast products being tested for interoperability, which is done much the same way all Wi-Fi products are tested for interoperability by the Wi-Fi Alliance. In fact, the lab doing the Gfast testing, the University of New Hampshire’s InterOperability Laboratory (UNH-IOL), also does Wi-Fi Alliance testing.
Mersh notes that one of the benefits of a certification program and beta testing is to drive things to market quickly, which is a challenge for every new tech. “Certification also gives you the end result of interoperability,” he says, which lowers costs for all. “The sooner you get to interoperability, the better.”
The Broadband Forum is also doing a lot of work on the software management side of Gfast, including creating an open access management platform that would work for many providers, and would have Gfast as just part of the specification. It’s in its infancy now, but as deployments grow, expect that to help the ISPs embracing the tech.
“I think the footprint [for Gfast] in the United States could be quite significant, in particular this MDU play,” says Mersh. “The number could be huge. I live in Atlanta, and the number of MDUs that Google and AT&T are after is large. Google is going more with fiber to the home. But if AT&T can do it without replacing existing wiring, that’s a huge benefit.”