The idea here is to provide very high speed from the street (which has fiber) to the home or office using existing copper - the last 30 meters. Laying fiber in the street is far easier than running it through buildings where every building has it's own challenges.
That's more or less what Verizon FiOS and AT&T U-Verse do, since many connections are still copper at some point on transmission to homes/businesses. There's some latency overhead I noticed in comparison to straight DSL or Cable though.
That's what BT Infinity is in the UK. I currently get 80Mbps down and 19Mbps up - which is perfect for offsite backups of family photos etc.
It's also totally unlimited, with no FUP - I've pulled down many hundreds of gigabytes per month (I work with very large data files and often work from home) and not heard so much as a peep from them about it.
BT infinity is using VDSL. This scheme is G.fast, which is basically to move the fibre even closer to the home and get more bits over the last little piece of copper.
Specifically, that's BT Infinity 2 -- confusingly, BT use Infinity for their fibre-to-the-home product as well, Infinity 3 offers 200Mbps and Infinity 4 offers 300Mbps. Availability is limited to a handful of exchanges, however
For more information, this appears to be attacking the "Cost to Connect" portion of the setup costs, which represent about half the cost of a FTTH deployment.
"In contrast, XG-FAST uses an increased frequency range up to 500 MHz to achieve higher speeds but over shorter distances. Bell Labs achieved 1 Gbps symmetrical over 70 meters on a single copper pair. 10 Gbps was achieved over a distance of 30 meters by using two pairs of lines (a technique known as “bonding”). Both tests used standard copper cable provided by a European operator."
30 meters with bonded pair for 10 Gbps. While this is pretty cool, due to distance limitation we are not getting telephone line modems that can outstrip fiber optics.
> While this is pretty cool, due to distance limitation we are not getting telephone line modems that can outstrip fiber optics.
From the article:
"XG-FAST can help operators accelerate FTTH deployments, taking fiber very close to customers without the major expense and delays associated with entering every home."
So you actually are getting a modem, hooked up to your telephone line, that goes up to 70m over telephone cable to the street, and then fiber the rest of the way. Makes a lot more sense than literally doing fiber to every home.
> So you actually are getting a modem, hooked up to your telephone line, that goes up to 70m over telephone cable to the street, and then fiber the rest of the way. Makes a lot more sense than literally doing fiber to every home.
This is exactly how AT&T Uverse works in brownfield retrofits.
I've seen this kind of setup in Japan more a few years ago, with vsdl as the technology for "the last meters": fiber optic was used by service providers down to neighborhood/building concentrators, and then the signal was carried over phone copper lines down to the office. We had a vdsl modem there, instead of a fiber one (I've seen those as well for business and homes).
Afair, we had close to 100Mbit/s on our lines. The building wasn't very old, apparently we had good copper wires (I can't comment on the average quality of those across the country, but knowing Japan, it should be good).
I'm not sure how distances like 70 and 30 meters are actually useful. In my home there is certainly more than 30 meters of cable between phone socket and phone company's junction box in the basement, add another 10 meters from this junction box to where the cable actually enters the building and then at least 20 to actually get to the street.
This is for Fibre to the Curb, Fibre to the Cab (junction box ) is what is more commonly used and you can already get plenty fast enough bandwidth for consumer's with VDSL or SDSL at that range.
Btw Cab is UK tlco slang for it not sure what the USA Telco term is
How is fiber to the curb that helpful though? Why not just run it up and put the box on the side of the house like they currently do? This requires an extra modem.
Hardly. You are forgetting that most people's copper cabling is degrading quite rapidly or is already of very poor quality. Furthermore the number of nodes required on the street due to the 70m limitation would be a lot higher than with FTTP.
Running cable on poles or pulling it through conduit is much, much, much cheaper than fiddling around in someone's attic or crawlspace. The appointment has to be scheduled (i.e., the homeowner has to be there), you have to worry about the three hundred pounds of boxes they've got stored in the way, the crawl space or attic may be brutally hot or cold, have wasps, raw sewage, mud, or any number of other hazards. It's a slow and expensive job.
Not to mention legal hooplah and potential billing. Digging up the streets requires permission from the city, and maybe you send them a bill for some of your services.
Digging up people's front lawns requires permission from every homeowner, and if you are offloading cost you have to bill each homeowner...
You have to have permission from the home owner to switch them over from ADSL to VDSL and to give them a new modem. Here in Australia the plan was to run Fibre to the outside wall termination point on each person's house by default (opt out). Then someone comes shortly after to patch it through the wall etc if and when the customer wants it. There was no legal hooplah at all.
FTTdp is the technical term. It makes a lot of sense, especially if the telco rolling it out has ownership and maintenance of both the copper and fibre end.
But none of this would make sense or is in any way indicative of the pile of shit the coalition is set to be responsible for here in Australia, lest anyone think that.
> While this is pretty cool, due to distance limitation we are not getting telephone line modems that can outstrip fiber optics.
I doubt anyone is under any illusions that the copper is going to outstrip fiber. However, it is still an important result: There are many places that are stuck with old-fashioned twisted pair and it takes literally an act of gov't to lay down fiber. Rather than leaving that population out in the cold, this result gives hope that they may have cutting edge connectivity.
The real issue is commercializability, which has not been Bell's traditional strength.
It applies to any call outside of the local loop. From the dsl wikipedia entry:
Because DSL operates above the 3.4 kHz voice limit, it cannot pass through a load coil. Load coils are, in essence, filters that block out any non-voice frequency. They are commonly set at regular intervals in lines placed only for POTS service. A DSL signal cannot pass through a properly installed and working load coil, while voice service cannot be maintained past a certain distance without such coils. Therefore, some areas that are within range for DSL service are disqualified from eligibility because of load coil placement. Because of this, phone companies endeavor to remove load coils on copper loops that can operate without them, and by conditioning other lines to avoid them through the use of fiber to the neighborhood or node
Anyone know if this would be feasible in Australia? Our National Broadband Network was originally planned to be fiber-to-the-home, but changed to a fiber-to-the-node system with the change of government. Not sure how close the average home is to its nearest node, so I don't know if it'd be less than 30 metres in most cases...
This is for fiber to the premise, not fiber to the node. It would have solved some of the problems with the old NBN model, especially in highrise buildings in the city.
I'm simply for this by the fact the government has been progressively removing all technical references about the NBN from every government website. During their election they promised a climate controlled box every 500 meters or so. This was ridiculed as completely out of the question due to cost and maintenance and has been quietly dropped since then. So even at the best case scenario the distance will be x20 as long as in these trials. And as the researchers noted: "giving broadband speeds up to 500 Mbps over a distance of 100 meters. In contrast, XG-FAST uses an increased frequency range up to 500 MHz to achieve higher speeds but over shorter distances."
You don't rip out the building because you're using the existing copper that runs inside the building. Maybe you'd run fibre from the junction box to the building then, or maybe you'd just use the existing copper on that route too.
Please read the article. This technology is only viable for copper wires shorter than 30m. The previous spec at 1/10 speed was for wire shorter than 100m.
"a major business benefit in locations where it is not physically, economically or aesthetically viable to lay new fiber cables all the way into residences. Instead, fiber can be brought to the curbside, wall or basement of a building and the existing copper network used for the final few meters."
I interpreted "the premise" as the building, but perhaps the conventional interpretation is the entire grounds. Either way, this is certainly about squeezing more out of the existing copper.
Have you seen the state of Telstra copper? No way, I can't even break 0.8MB/s with ADSL2. The pit outside my house is filled to the brim with water when it rains, which drops the sync speed even further. The MDF of my building was hanging from the wall by the copper lines, which Telstra has deemed to not be an issue. Heck, the lines are just twisted pair buried straight in the ground with no protection at all.
Copper can take quite a bit of abuse, the real issue you will find on your line will either be in the pit (Telstra trialled a program with a anti-water gel which had a reverse effect and corroded the copper connections). Or you have the issue of wiring within your house, even though household wiring is typically straightforward most houses in certain areas will have Optus and Telstra copper running into the house, how those are wired into the actual wall socket is another story as Telstra are only required to supply a working phone line if you sign up for TPG/Dodo or any other ISP that isn't Telstra. Sign up for Telstra and they ensure all wiring is nice and only distance/exchange conditions are a factor.
I redid the wiring in the house as far as I could and cleaned up the MDF myself (screwing it back on the wall was a nice start). Problem is probably in the pit, but as you said it's not their interest to fix it as I don't have a Telstra service at all.
Replaced the end points, not the wiring itself. The wiring is 2000s vintage and not anywhere near as terrible as the huge old Australian baseboard plugs it was fastened in to. Replacing the wiring in the walls was a bit too much given that I expected to be ditching it for fibre soon.
It would be. This wouldn't however involve the nodes as built at the moment. The problem is one of distance, not of the protocol that you're using. Adopting this instead of VDSL2 and keeping the architecture and topology the same won't make any difference worth speaking of to most premises.
Instead the premises will have in the pits (or on one of the electricity poles) just outside them a tiny box - powered from the household itself through the existing copper - that connects to the fibre, which runs past the household. This contrasts with the average distance to the nodes of about three, four hundred metres likely under what the coalition is doing.
It gets rid of the most expensive part of a FTTH install, which is the actual connection to the customer. So, yes, it's viable.
It's also, however, substantially more like Labor's plan than the coalition's. So even if it were to make a buttload of sense and be insanely more cost effective it still wouldn't be done because Labor is evil. Or something. And Telstra and Murdoch wouldn't approve of something like this.
It would make the installation process easier, in that there would be no need to pull fibre through existing lead in conduits (which is a major issue) This could be called 'Fibre-To-The-Fence' basically.
I bet this isn't that much cheaper than a full FTTH - the cost advantage of DSL is eroded as the port density goes down.
Although this is spun to sound like a breakthough, it's basically a cost reduction measure. G.fast and this new XG.fast are there to avoid the hassle of negotiating with each householder as to how new fibre gets deployed on their property, when it's a lot easier to negotiate with the municipality all in one go.
The greatest breakthroughs provide something for the masses at a lower cost than previously possible. An important part of engineering is leveraging existing infrastructure to improve service quality. As an example of this it has been technically possible to put fibre into every home for years but apparently not feasible, price-wise. Speed is speed whether it's copper or fibre.
"speed of 10 Gbps for transmission of data over traditional copper telephone lines"
Good news for some places where they still use copper telephone lines connect to the Internet. But in my view, Chinese operators don't like this technology, they more like optical fiber. The first, it's faster than copper telephone and the future copper telephone lines will completely replaced. The second, in countrysides some bad guys steal the copper to sell for money, it's a big problem for operators.
Take for example remote/rural communities however. It is enormous investment to run fiber out to these communities. Eventually, the copper will need to be upgraded... In the mean time, this is an excellent stopgap measure.
In some ways, you are right. In China, some facts prove countryside's copper is often stolen.Chinese operators has a same measure, they call it "Fiber In, Copper Out"(translate for English it's so hard for me), first is improve network speed, second is reduce copper costs due to stolen. They prefer build fiber in countryside, rather than upgraded copper.
I worked for a consulting enginer in the 80's and our telecom guy said the same about African states use microwave for long links as the copper gets nicked.
What is much more interesting here is G.Fast, the technology it is based on.
G.Fast, is suppose to be a FTTdp, where fibre goes all the way to your building and replacing the central copper wiring box with its DSLAM. In countries where you have many flats within a building, it is likely the copper wiring to home socket are all within 100M range.
I am not sure this applies to every scenario in every countries. But the concern of Copper quality is mainly a problem with copper outside of the building. Copper within the building are in much better shape and form.
G.Fast also allow operator to config the upload and download speed. Hopefully that means at least 100Mbps of upload speed in the age of cloud computing.
I cant wait to see this deploy. No longer would teleco needs permission and planning to work around every home and every building to get high speed internet.
Huawei and Alcatel-Lucent are the fore front in G.Fast technology.
This is interesting, but as has been pointed out, the distances are very limited. Just something to think about, 802.11ac comes pretty close to the same speed with no wires at all, so, it kind of makes you think, why even bother with the wires at all.
802.11ac does an order of magnitude less speed than this inside an RF isolation box. Bring it out into the actual world with other people using it, and you aren't getting anything close to a gigabit.
The problem will be getting regulatory approval. A high speed digital signal at VHF is going to radiate like blazes and wipe out FM Broadcast and hundreds of emergency services.
This system will face the same problems as BPL, only much, much worse.
The truth is that you can't reuse the (all ready full) radio spectrum, especially over copper.
The only way you can do it is with well shielded Coaxial Cable (eg Cable TV) or via light (eg Optical Fiber).
It's a triumph of Sales people over Engineering. Or to put it another way, the laws of physics will always win.
If Comcast said, "I can get you a 10Gbit connection, it's fiber all the way, but it's going to end as an SFP port at the pole/wiring closet, and you have to take it from there."
I would have thought homeowners and businessowners alike would spend the money to upgrade their own "infra" to make that work, rather than something like this. Obviously AL thinks there's a market here, it just seems like such a waste.
Wiring isn't under control of the home/business owner until it hits their NID, which is already on-prem. Getting the connection to the NID is where the expense is, and it's why AL is putting money into the process. Without this, someone has to spend the thousands of dollars pulling permits/digging trenches to run fiber to your house.
I guess I don't understand why you need this much tech just to get the fiber off the pole and into my house [1].
The coax drop from Comcast drapes off the pole, hits little more than a barrel connector on the side of the house, and then comes straight in the house and plugs into the modem.
30 meters seems pretty useless, even for a FTTN setup, it would surely be simpler just to run fibre all the way then build nodes every 60 meters or so.
If you're on 1.4 Mbps then the odds are very good you're too far from your exchange for this to make even the slightest bit of a difference.
Not just by a small factor, but at that speed you're likely to be around two orders of magnitude too far away for this be in any way useful due to the attenuation inherent in your copper line.
The only way this latest increment is useful is to bring fibre in your case about 99% of the way to your household. This will bridge the last 1% because it's cheaper than running fibre into your household.
7/10/2114
Alcatel-Lucent-Comcast-Timewarner megacorp sets record of 10 Tbps over 200 year old copper telephone lines. CEO states "We'd really like to bring you broadband but we've been fighting this last mile issue for quit some time. This new breakthrough will finally solve that."
As some others have mentioned, this isn't about getting internet from the exchange to the home/business, just from the street. In this context 30 metres should be ok for the vast majority. From the article
> It will enable operators to provide Internet connection speeds that are indistinguishable from fiber-to-the-home services, a major business benefit in locations where it is not physically, economically or aesthetically viable to lay new fiber cables all the way into residences. Instead, fiber can be brought to the curbside, wall or basement of a building and the existing copper network used for the final few meters.
Because every household has copper phone cables run to it already. You can run fiber to the trunk and have high speed communication using existing copper lines.
> XG-FAST can help operators accelerate FTTH deployments, taking fiber very close to customers without the major expense and delays associated with entering every home
How many of those existing copper cables are of high enough quality that they can support anywhere near that speed though? The copper lines in my parents' house are so bad they can't even support a second phone line, much less decent networking speeds.
You can dig up an alley and re-string from poles with little hassle.
Digging up everyone's yard, replacing lines that might now be running under garages or driveways, and just generally having to operate in people's yard is a lot more work. Especially multiplied by tens or hundreds of thousands of homes.
If you can just trench an alley and run fiber to the nodes then re-use the existing copper for the last 100m that's probably a huge cost savings.
It doesn't require sending out a tech to pull cable through random crawlspaces, for one. If the tech can hook everyone up at the pedestal (or whatever) without having to run cable into the individual houses it will save a ton of money.
Hey, I wasn't joking- well, not that much. You will get more disputes when you deploy across someone else's property, especially if you break or uglify something. That's a serious reason for stopping on public land.
Nice technology, but I have had it with telecom company excuses. Electricity has been deployed to most homes in rural areas, as well as copper telephone. It should be no different with fiber. Why could it be done then, but not now?
Didn't telecom get billions of tax $ last decade for broadband deployment, with nothing to show for it?
Anyway, looks like they were interest free loans vs. outright subsidies. At least something came out of the New Deal, whereas fedgov spending during this recession is pissed away.
Someone correct me if I'm wrong, but I believe they got tax breaks (tax incentives). The numbers thrown around have tended to be intentionally over-estimated by a huge factor, typically critics will proclaim the telecoms got $200x billion, the real number was closer to the low tens of billions (substantial, but not $200b crazy).
Such a shame, and a sham. Pay for internet to schools, and you get a single machines hooked up to redundant Cisco edge routers capable of serving 10,000 nodes with ISDN uplinks for $1,000/mo.
On the other hand, give fast track regulatory approval for poles and right-of-ways to ISPs for fiber infra and you can wire a whole town, and let the library pay for their 100Mbps line like everyone else. Bet you that library ends up with a properly sized $100 SOHO router and a $100/mo broadband plan if they have to pay for it out of their books budget.
I think at least one side of the equation or 'market' has to be driven by someone who is paying with scarce money, not just Benjamins hot off the presses, to keep things under control.
Whatever the theoretical limit is, they're pretty much right at it at this point. Similar to how ADSL2+ was near enough at the theoretical limit of what the frequency spectrum it used could do given the attenuation it was asked to do it at.
For the same frequency ranges as ADSL2+, VDSL2 delivered pretty much no improvement. What VDSL2 did do is expand the frequency range up to 17 MHz/30 MHz, depending on the profile.
So the question you might want to ask is "How high a frequency can we use?" and in that case the real question you're looking for is "How short a piece of copper do you have?".
And the answer to that is very quickly becoming "a very short piece of copper indeed".
Either that or a lot of copper that's ideally also very thick... which, considering most copper was rolled out with the aim of providing a low attenuation at 3000 Hz over at most a low number of kilometres isn't often the case.
The bandwidth of an electrical link is a function of bandwidth (the range of frequencies present on the cable) and the encoding (bits per second per hertz). The first on is limited by attenuation on the cable. As you increase the frequency, the cable does a less good job of propagating the signal, and your signal level decreases at the other end. This is limited by power and cable quality. The efficiency of the encoding is limited by the sensitivity of the receiver and its internal noise along with external noise.
This system uses a bandwidth of 500 MHz over two pairs. That gives roughly 10 bits per second per hertz. Both are pretty close to the state of the art over unshielded copper cable. You coaxial cable is usually about 600 MHz. But if you use waveguides and the like, you can get into the GHz range.
Depends on a number of factors: the major one is the length of the wire, but also what gauge of wire, how well twisted it is (phone wires aren't very twisted due to not expecting to use the high frequencies), the radio environment, local radio interference standards (telephone wires are leaky at these frequencies), how many of your neighbours also have DSL (although this is improving due to crosstalk cancellation mechanisms in the newer standards), how many watts the local operator is prepared to put down the wire...
I just want some competition in high speed connectivity. I don't care if it's fiber, copper, wireless, coax, power line, or sewer. If this makes it cheaper for me to get 100 Mbps or higher, sounds good to me.
Actually, this invention requires Fibre to the Curb. The only section that would remain copper is the curb to wall-socket link.
And in the long term, you'd probably pay to replace this yourself with fibre since (if other high bitrate copper solutions are any indication) fibre would be lower power, lower maintenance and more reliable.
So what, that changes absolutely nothing about the monopolistic system of our infrastructure. All this will lead to is hidden collusion on the part of the providers just like is done in the mobile phone and data sector.
We have a failed government with incompetent and neutered regulators and legislators that are operating in a system that is so deeply corrupt that the best of intentions cannot generate any kind of real momentum. There really needs to be a focus on bifurcating the infrastructure from services. What is wrong with us that we pay taxes to pay for functions of government when all that happens is that those taxes go to subsidizing the privatization of government services and then being overcharged by that very private sector we stood up.
Infrastructure is a core function and purpose of government, without it being under socialized responsibility and control there is no purpose for government.
Another example that is so blatantly obvious but no one seems to see for all the trees is electronic payment processing; why is there no almost zero cost electronic payment infrastructure that essentially serves the same purpose as paper money. I am not at all for the abolition of paper money, especially in light of the reminder that government is not our friend or anyone's friend but its own, but there should be a government / public infrastructure for payment processing. It's actually rather shocking that there is no such system and it is purely corruption of the worst kind that is preventing it.
Ever wonder why it is rather difficult to use a Visa/MC/AE in Germany, it's because the electronic payment / money system is a pseudo public infrastructure and highly regulated. There is no money to be creamed and siphoned off of the economy by the big payment processors.
Something the startup community does not take into account is that it is only as successful and has as much wiggle room as the big cartel controlled sectors permit. So you built a service or a product, you're just mucking around while you still have to pay protection money to the mob.
Sometimes I wish that efforts like Dwolla would gain a lot more traction. Their payment processing fee is fractions of the extortion of the mob, yet they can't seem to catch on. What is wrong with us, why would we be ok with having your profits siphoned off? Those huge payment processing fees are profits that you pay for not good reason.
The same problem affects this technology, there is absolutely no reason for any ISP to implement this technology any time soon. They get away with extortion for shitty services, why would they provide exponentially faster service for anything but a lot more as they seesaw up expectation that "of course, it's faster so I should pay more"
How many of you guys actually hook up any of your equipment to a hard line? I've used nothing but wifi for years. I'd be more interested in ultra-high speed wireless that could stretch to 30m and get me to that fiber node.
Wifi is very convenient, but adds latency, drops more packets than a CAT5 cable will, is (theoretically speaking) easier to eavesdrop upon, and degrades much more quickly than wired connections do when the number of clients is high. That's enough for me to prefer wired ethernet when possible in the present, and those problems would only get worse if you're sharing an access point that is outside your home/apartment, that you don't control, with dozens of people. Plus, I'm no RF-engineer, but I think the "high bandwidth" and "relatively long distance transmissions that must pierce thick walls" goals are at odds with each other, especially when working at wifi's high frequency range.
It's possible, but then the Wi-Fi router will be outside of your home, probably mounted on the poles, and owned/operated by your ISP. Who will be happy to charge you by the number of devices you can connect.
No thanks, I would much rather have direct control over my own home network. (Sure, you could achieve that with a repeater in your home, but then what was the point of moving the Wi-Fi router to the street?)
My home PC is hooked up to my router with a CAT6 cable. The router is hooked up to the FTTH modem with another CAT6 cable. So it's hard lines all the way. Both the speed and latency of this setup are much better than anything I ever got with any wireless alternative, especially when I'm trying to talk with other devices on the other side of a wall or two.
AT&T / Verizon (and the cable companies) will never give up their golden goose unless threatened by competition. They can do FTTC all day long - unless they actually uncap or otherwise significantly change the subscriber bandwidth it means nothing to the end-user.
Nice intro to the theory around data transmission limits: http://newsoffice.mit.edu/2010/explained-shannon-0115