In building deployments would mostly be below Ethernet cable run limits, so why use this tech? In building it would make more sense to stick with Ethernet.
The long-awaited G.fast physical layer standard has been signed off by the International Telecommunications Union (ITU), improving the chance that the world will see standards-compliant, interoperable kit start to ship in 2015. The ITU almost manages to work up some excitement about the announcement, reminding a grateful world …
Monday 8th December 2014 09:32 GMT Terje
Wouldn't it be a more fruitful pursuit to work on increasing the reach of lower ~100Mb tech so as to be able to reach more households with decent speed, each new step seems to be just faster with even more limited reach. This tech now seems like an extreme niche application to me bordering on a waste of money even putting it on paper..
Monday 8th December 2014 10:06 GMT Mage
Probably there isn't much more reach possible at 100Mbps. So they experiment with shorter distances which inherently allows more speed.
The future is fibre. Even VDSL2 is becoming a Niche for existing apartment blocks, hotels, guest houses etc that don't have Cat5e or similar. At about 900m to 1200m (depending on cross talk, noise, cable condition etc) the ADSL2+ becomes better than VDSL2. Most VDSL gear will fall back to ADSL2+ on poorer / longer lines.
Another aspect also is RFI production as you increase power and bandwidth for higher speeds.
Monday 8th December 2014 10:09 GMT HMB
Monday 8th December 2014 10:38 GMT Terje
It's not that I'm not aware of the physics wall it's just that I just don't think that we should be even close to that wall for the applications we are talking about. While of course the largest issue is the quality of the copper there has been so much increase in speed and "smart" solutions to problems that I just fail to see that for the last 10-15 years we are still stuck with more or less the same 24 Mbit max offerings unless you live "on top" of a tele station. What other communication standard is still the best available after such a long time?
Tuesday 9th December 2014 15:49 GMT Alan Brown
Monday 8th December 2014 10:07 GMT HMB
It's Worse Than That It's Physics Jim
Probably best to complain to the person who invented physics on that one. Bloody inverse square law :P
I'd encourage reading up on this to understand why developments have followed the course that they have.
However, my quips aside, while there's only so much that can be done about boosting signal over line length in ideal theoretical conditions, in reality much interesting work has been done on compensating for cross talk with a technology known as vectoring. That's probably the best you'll get for your request. In the mean time you could always lobby for increased signal power and spend your life dodging angry hoards of radio hams. Don't forget though, every time you pump twice the energy into the phone line you'll get a 3 dB increase in signal and you quickly hit a situation of diminishing returns.
As an alternative you could push for line bonding, but if BT have to go out and install an extra new line for everyone, the economics of it quickly go down the toilet, you might as well just get them to roll out fibre.
Monday 8th December 2014 10:09 GMT Andy The Hat
Monday 8th December 2014 12:54 GMT Jellied Eel
As others have pointed out, a large part of the market are things like shopping centres, hotels, apartment blocks etc etc. There, it's about potential to re-use existing cabling, cabling distance, subscriber density, contention and all that good stuff. Plus avoiding OAM costs.
So wiring up Ethernet would mean installing additional ducts, cabling, wiring closets. Costs may vary depending on property type and building regs. So depending on building size, closet may need an Ethernet switch trunked back to one or preferably more main switches. Then each subscriber would need their own VLAN and some way to be able to point that VLAN at their chosen provider.. Unless you're not intending to give them any choice. And then it becomes a question of who'll manage and support that lot, especially if they're trying to manage multiple properties.
Then the ability to slap on a G.fast modem and having it work much the same way as a regular xDSL subscriber starts to look attractive. It's also a way to improve on existing pseudo-FTTH deliveries where the 'F' is really copper and the ASA lets ISPs get away with it. Where properties are dense enough, G.fast can aggregate subscriber connections back to a street cab & fibre conversion can happen there.
Monday 8th December 2014 12:51 GMT Well Known Cowherd
There might be alternate uses of this tech outside of connecting houses to the interweb.
I used vdsl2 for connecting two buildings via a much too long run of cat5e. I could see a similar situation where someone needs to connect a couple of closets together using existing (terrible) infrastructure.
Tuesday 9th December 2014 15:48 GMT Alan Brown
In other news
It's cheaper to run fibre to the house and it'll go longer distances than trying to force Gb down voice-grade twisted pair.
Even if Gb over voice worked at slightly longer distances the odds are pretty good that it'd turn to shit as soon as there were more than 1 or 2 active circuits in a cable, for the same reasons that VDSL does.
The same thing would happen in a "dense environment" (the shopping centres,. etc etc mentioned, at much higher cost than just pulling an ethernet cable )
This has become a solution looking for a problem. Unless these things are small enough to fit in a street jointing box AND be powered via 48V @50-100mA over twisted pair AND be fed by fibre, they simply aren't viable - and if you have to run fibre up the street to feed heads spaced every 50 metres you might as well run it into the customer premises at a lower overall cost.
If they can scrap copper as they go (and not have thieving contractors making off with it), openreach might even be able to turn a profit as well as making the network less attractive to midnight cable rats
Wednesday 17th December 2014 05:40 GMT Medixstiff