Slashdot Mirror
ITU Standardizes 1Gbps Over Copper, But Services Won't Come Until 2015
alphadogg writes "The ITU has taken a big step in the standardization of G.fast, a broadband technology capable of achieving download speeds of up to 1Gbps over copper telephone wire. The death of copper and the ascent of fiber has long been discussed. However, the cost of rolling out fiber is still too high for many operators that instead want to upgrade their existing copper networks. So there is still a need for technologies that can complement fiber, including VDSL2 and G.fast. Higher speeds are needed for applications such as 4K streaming, IPTV, cloud-based storage, and communication via HD video, ITU said." Meanwhile, I'm hoping Google Fiber, FIOS, and other fast optical options scare more ISPs into action along both price and speed axes.
I'd be happy with 10 Mbps since I'm stuck with 3 Mbps from CenturyLink unless I want to do a deal with the devil... err, I mean Comcast Small Business... and have them dig a trench for the cable to my house through the yard and landscaping.
Why would FIOS scare Verizon DSL into action?
This still won't fix the problem of some ISPs having a monopoly over some areas, such as Télébec in small Québec regions.
Get free satoshi (Bitcoin) and Dogecoins
DAMN... at least once every 10 years pick a broadband solution and BUILD IT ALL THE WAY OUT. To every last house in the US. This never ending cycle of new technology coming out and being bult out to the edges of the big cities and then the next new technology hits and they stop where they are go back to the center of the big cities and start building out again.
Just once. Get something other than dialup and satellite all the way out to every last house in the US.
Digital is, by definition, imperfect. Analog is the way to go.
...will again become the host that one is connecting to and that side of the network getting there, rather than one's last-mile issues.
As for FIOS, I can't deny that I worry, ever-so-slightly, about security. From what I understand, there's a single fiber pair that feeds numerous subscribers, and there's some kind of fan-out kit that sends the same signal from the service provider to all of the subscribers, and that phone company active equipment on each customer premises filters out all but the traffic intended for that subscriber. My concern is that if the phone company's gear gets hacked by either another subscriber in the segment then they'll be able to get at traffic that they shouldn't.
Do not look into laser with remaining eye.
no one will hardware their way out of a problem. Lowest cost possible, with the savings passed on to the customer. Yeah right.
I'm stuck in copper-land thanks to the phone monopoly in my town, and the copper we have can't reliably transfer data at faster than 8Mbps. 15Mbps was great when it worked, but the disconnects were frequent. The residents in my town are never going to see gigabit speeds over our copper infrastructure. The phone company has no reason to improve it. There is no fiber alternative, Verizon pulled out of our state. Our cable TV monopoly is equally disinterested in provided higher speed service. This is probably a significant challenge all over the United States. We need to find a way to revive competition and get these legally-sole-provider-in-the-region companies to offer improved service.
DirecTV forced cable companies to up their HD offerings by making over a hundred channels HD in one go after launching some new satellites. Before that, none of the cable MSOs would bother. We need a similar antagonist in the ISP space.
Meanwhile, I'm hoping Google Fiber, FIOS, and other fast optical options scare more ISPs into action along both price and speed axes.
Why would they move along an axis that significantly reduces profits or increases costs, when they can continue to throw legal caltrops under the wheels of progress?
There's room for argument over how expensive it would be to buy more backhaul capacity or reduce subscription fees, but there's little doubt that buying utility commissions and legislators is a lot cheaper.
OK. So long as G.fast is an improvement over what they're using now, that's a good thing. But until/unless I can get 1 gbps at my desktop, I don't think they should be allowed to advertise it as "Gigabit Internet."
This is the typical phone company thing... "buy Internet service from us!" How fast will it be at my house? "Um, we have no idea!"
"1 Gb/s over copper" is something that's existed for a looong time.
1 Gb/s over a single crap twisted pair copper on the other hand...
and we only get .8MB speed - and i'm grateful for that (although i have seen connections in the 1Mbps range)
the Verizon on-line service tool said 'service not available,' but a little social engineering with the local business office got us the service
Crosstalk alone I would think would be an issue. Pumping that kind of data, even if it is digital, the amplitude on the line would introduce crosstalk I would think. Granted, it is analog, but a 33.6 fax over copper causes headaches. I just wish people would give up the fax machines, and use secure scan to email. Trying to get a V.34 modem to work on a VoIP line is a headache since most network guys (for obvious reasons) locate the ATA box in the equipment room, by the time you string a RJ11 cord all the way to the fax, the signal level has dropped to the point that there isn't enough current on the ring signal to trip the relay to tell the fax to answer the line. Then, introduce a little attenuation into the line, after the handshake signal, and you end up with it doing multiple retrainings trying to get a connection fast enough to send a document. I just wish fax machines would DIE DIE DIE.
Then it won't come until... ever.
Since it only work for 100 meters, you're just as well off putting in a Cat-6 or multimode fiber pair and run 1000B-T or 1000B-SX between the points and skip the extra CPE and other equipment...
My telecom just rolled out 40Mbit service, about 10 years after Comcast did the same. I don't expect to see this any time soon unless Comcast uses it (they're Fiber-to-the-Neighborhood and then copper, so it's possible). I still won't do business with Comcast, even if I can basically make pricing a wash with bundling. I also could get a DirecTV bundle but giving up DISH would be hard, plus I don't give a rip about sports, which is kind of the focus of DirecTV.
Oh, it will, but not in the way you want.
---- Booth was a patriot ----
Even as current tech slowly inches faster, our caps are going down.. So we get 1g to the house which is cool, but use our monthly allocation of data even faster, which is not cool.
---- Booth was a patriot ----
Prior to divestiture, AT&T would replace the copper loop between the CO and home every 25 years. Post divestiture, SBC ( Masquerading as AT&T ) will not replace copper until it rots into the ground. When it does get around to replacing it, the loop runs to an RT or equivalent instead of the CO. One of the biggest excuses for the substantial increases in your phone bill was for maintenance of the local loop. They have had more than adequate time to replace copper with fiber, but have chosen to pocket the difference instead. Now, after taking all the money for the maintenance that was not done, they want us to pay again for the new loops they will inevitably have to install.
"To those who are overly cautious, everything is impossible. "
if G.fast distance from CO office to home is 100 meters, then good luck - we're not going with another DSL fiasco installation again.
Ya no... that's not how it works at all.
It's called a Fiber Mux: http://en.wikipedia.org/wiki/Multiplexing
That's how things like DSLAMs work: One (or more for redundancy) fat pipe for backhaul, a router or switch in the box at the curb, and individual links carrying only each customers' data to the DSL modem at each customers' site.
Passive Optical Networks work like cable internet (and vaugely like the original party-line coaxial Ethernet): A pair of light frequencies (one outgoing, one incoming) connect the box at the curb, through attenuator/splitters, to each of a handful of sites. (The one I saw had an 8-bit hardware address and handled 250 subscribers per fiber.)
- The outgoing signal contains the traffic for all up-to-250 subscribers on a given link and the subscriber box rejects traffic for all but its own destination(s).
- The incoming traffic takes turns on the other light frequency. (Timing information is on the outgoing link and they run a link-level protocol to assign slots as requested when the subscribers' boxes have traffic, rather than a collision/retry protocol.)
Advantage is you need about half as many optic transcievers to implement it, while optical splitter/combiners are really cheap.
So, yes, it would be trivial to build a box that could listen to the fiber and tap your neighbors' downbound traffic. (You MIGHT be able to tap the upbound traffic from SOME of your neighbors, too, with a sufficiently sensitive optic receiver and if the fiber joints and splitter/combiners have enough discontinuity to reflect enough of their inbound light.)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Crosstalk alone I would think would be an issue.
Yes, it is. The standard is largely about canceling crosstalk. (Look for "vectoring" in TFA.)
Without the standard's crosstalk cancellation feature, but with everything else according to the standard, the speed drops by a factor of five.
200Mbps over these short hauls is not to sneeze at. But it's not such a big deal, either.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Why don't they call it Gigabit DSL? Now it's G.fast. Before it was "Very high speed" DSL. What's next? G.super-mega-ultra-fast?
Really, in Soviet Russia there are enough poles that don't belong to phone company, be it the trolleybus or tram poles, lighting poles, roofs of buildings a.s.o. (The phone network is almost totally monopolized by Rostelecom state monopoly and is underground buried in asbestos-cement tubes almost everywhere). The alternative providers use these poles.
The longer they cling on copper wires, the further upload speeds fall behind. IMHO, this holds back a lot of data intensive innovations and slows general development of the IT sector.
Someday, this is going to seem unbearably slow, and whoever decided to put the word "Fast" in the name is going to seem like an idiot.
A pair of light frequencies (one outgoing, one incoming)
I don't believe! It will mean that either there are 250 lasers of different color and some "prism" with 250 outputs to separate and combine them all, or 250 precisely tunable lasers and the same "prism". The TDMA scheme where lasers are ON in their dedicated time slots looks much more affordable.
I understand that such methods are applied in some deep sea cables but...
Now, my Internet link is a Cat-5 FTP hanging on a steel cable. If some day the thieves will cut it off (which is quite strange since the territory is guarded), I'll install a P-274 Soviet Military Field Phone Cable. It's dirt cheap, strong as hell and has inseparable copper and steel braids. No scrap traders buy it.
You're thinking of a different scheme: Wavelength division multiplexing. That would be about as expensive as separate fibers to each house with individual transcievers. (Moreso, since the many different-colored laser transcievers are pricey.) Wavelength division multiplexing is about getting more bandwidth or channels out of fibers, in long-haul or around large datacenters.
There aren't 250 frequencies in PON. There are only two. One for signals from the curbside box to ALL 250 houses, one for ALL the houses toward the curbside box. (The system I'm familiar with used the two common infrared laser diode frequences, which are far apart to easily filter.) They only use two colors so they don't have to worry about reflections from imperfect fiber joints and the like, and so the subscriber boxes don't have to sort out each others "talking", too.
Single transceiver in the curbside box for ALL the 250 subscriber sites. That's where the big savings from PON comes from. (There's a bit more by using one fiber for all and splitters, rather than separate fibers for each.)
The houses are sorted out by time division multiplexing, not separate color lasers.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
(Moreso, since the many different-colored laser transcievers are pricey.)
Modern fiber ONTs have programmable lasers. They can very selectively tune the wavelengths used, down to the nano-meter of the wavelength used. They can also tune the bandwidth, spread, and guard size. Even the shape of the wave form. All packed into a $150 chunk of technology.
I'm happy with my download speeds (I won't complain if its faster lol) but my uploads sucks.
That would be great if there weren’t any House Republicans. Our new conservative government in Australia has decided that the previous government’s plan to do that (to 96% of houses) was not a good idea, and instead we’ll all be getting 25Mbps by 2019. Good one!
I’m old enough to remember 16K of memory being described as “whopping”