-
I wasn't sure if it was right to jump on to the other G.fast threads for this.
I was looking around youtube for G.fast videos and noticed these. I haven't seen these before.
G.Fast Huntingdon
[youtube]https://www.youtube.com/watch?v=VIze5vKPXQM[/youtube]
Using G.Fast technology to deliver fibre-like speeds over copper
[youtube]https://www.youtube.com/watch?v=AWcb5TkftEU[/youtube]
G.fast trial customer feedback
[youtube]https://www.youtube.com/watch?v=dJYqDNnrZaE[/youtube]
-
Nice find.
While I'd love to get excited at the prospect of these videos and G.Fast as a technology, knowing that BT have decided to not install fibre to the DP is extremely disappointing.
For those who haven't heard, BT's current plan is to simply install G.FAST DSLAM's at the current FTTC sites. Nobody's copper/aluminium telephone line will be be any shorter.
As I understand it, if you current sync is less than 60Mbit (possibly 70mbit) on VDSL2, then you won't even be eligible to order G.fast, as your line will be too long.
-
Here is my graph before I had a ton of crosstalk, the relevance is the snr and bitloading on the end tones which gives an indication of how the line would be using the higher tones on g.fast.
These graphs are more relevant than current as g.fast has vectoring protection.
I currently sync at 71.5mbit, I think my line would not get a very good sync speed on g,fast (remember it will have power cutback for entire VDSl2 range).
note tho the bitloading on the end tones is lower than capability as it had a max sync with spare snrm so didnt need to max out the bitloading, 3snr per 1bit. So 45db snr needed for 15 bits. 36snr needed for max 12bits on g.fast. Also this was on my old pair, not the same one I am on now.
--edit--
These bitloading and snr is actually with "some" crosstalk, it was taken after my attainable dropped to 90 from 110, I didnt manage to make graphs whilst I had my original stats. So the line without crosstalk is better than this. Attaching QLN also.
-
Forum in Focus: G.fast
Following his presentation at Gigabit Copper, Broadband Forum CEO Robin Mersh spoke to Total Telecom about G.fast.
[youtube]https://www.youtube.com/watch?v=d4ca9SA90yo[/youtube]
Swansea G Fast Trial 08 02 16
[youtube]https://www.youtube.com/watch?v=vZco5qJUZsU[/youtube]
G.fast Solutions at Broadband World Forum
Calix vice president of product line leadership Shane Eleniak reports from the Broadband World Forum show floor on the latest developments in G.fast, including the Calix and Sckipio demonstration of G.fast bonding technology delivering broadband speeds in excess of 1 gigabit per second (1 Gbps) over existing copper infrastructure up to approximately 250 meters.
[youtube]https://www.youtube.com/watch?v=suZ7LI-9GMM[/youtube]
-
An old'ish video from October 2015. I thought it might be of interest and Peter Bell is speaking.
BT Updates on G fast Plans
Peter Bell, CIO at Openreach, the access network division at UK incumbent BT, provides an update on the operator's G.fast trials and how Openreach is planning to deploy the broadband technology in its street cabinets.
[youtube]https://www.youtube.com/watch?v=vJyCkRSIsYY[/youtube]
-
Yes well - let's hope it works better than G.Inp! ::)
-
How BT is pushing the development of G.fast broadband Video posted March 2016
UK telco BT is placing a lot of faith in the ability of G.fast technology to over-deliver on its original brief of very short range speed and capacity enhancements and extend its operational reach to distances up to 400 metres, enabling this node-based architecture to complement BT's existing VDSL2 rollout and GPON deployments. In the run up to the Broadband World Forum event in London next month, BT's MD of service strategy and operations, Mike Galvin, talks exclusively to TelecomTV about the status of its G.fast trials, its overall broadband strategy, and why he believes BT will see off the threat from Google for control of the connected home. Filmed at: BT Centre, London, September 2015.
[youtube]https://www.youtube.com/watch?v=eDclfhl8Ad0[/youtube]
{Moderator edited to fix a defect due to finger slippage.]
-
Hi
So from that video I got, G.fast originally designed for distances of 75 metres, "has been fiddled with" by BT to go to 300 metres and as far as 400 metres.
Basically what they are saying is a technology that was purposely designed to run at low power over a very short distance with little cross talk and radiated interference in order to provide much higher guaranteed minimum speeds (i.e. minimum of 300Mbps), will now be fiddled with to run over 400 metres with all the problems of DSLAMs having to mess with parameters, cross talk meaning your speeds only get slower as more people come online, vectoring problems, big holes in the frequencies used to avoid interference with FM radio, big chunks missing to avoid interference with ADSL and VDSL that run along side, just so a few lucky people close enough get a significant increase in speed (those within 75 metres that G.fast was designed for), and everyone else on the margins still trailing way behind with a few extra meg and a higher monthly bill.
Congratulations BT.
Regards
Phil
-
You had all of those problems anyway.
G.Fast was given speed targets at various distances from 50m to 250m (150Mb @ 250m, 200Mb @ 200m, 500Mb @ 100m, 500Mb @ 50m with VDSL2 frequencies left alone, and allowing a Max of 1Gb @ 0m). It was never /just/ a 75m system.
The ITU had scenarios from multiple sides, not just BT, in setting those targets. Those problems needed solving and including in the specs.
The G.Fast targets ultimately reflect the realistic dreams of 5-8 years ago - that, if utterly lucky, copper could do gigabit. Just. Over the shortest of distances. But an aggregate of 150Mb at 250m is no better than vectored VDSL ... you just wouldn't target a multi-billion rollout to go just as fast as the previous rollout - which was a consideration to BT, but not all operators. The numbers say that a DP deployment was the only way to offer a jump in speeds for BT - so they probably saw it at the time as a 75m system more than anything else.
Nowadays the dreams are of 5Gbps over the shortest distances, and G.Fast has provided a framework that is far more capable than they ever expected. The extra capability now allows a significant jump in speed to be offered at distances of 200-250m, and a reasonable jump at distances of 300m-400m.
You might think of BT as having fiddled with the spec. However, if anyone realised what G.Fast would turn out to be capable of, those changes would have been in the original spec. I think, however, the telcos realised there was value in having V1 of the specs approved early, and that the changes could be best dealt with as subsequent amendments.
BT are, of course, one of the main drivers in research for G.Fast, so we should expect the standard to reflect and include their own requirements. But they're not alone ... and you can get a flavour of the research process from a Dutch perspective by looking at some of the documents here:
http://www.joepeesoft.com/Public/DSL_Corner/_Index.html
The presentations on 4GBB are interesting, but one of the more interesting is the "KPN requirements for G.Fast" - which shows a lot of the thought processes within KPN and and TNO with regard to deployment scenarios, and distances.
Swisscom are another example of an operator at the bleeding edge - where they are deploying their FTTS concept, with a range of 200m in mind.
-
Hi
So from that video I got, G.fast originally designed for distances of 75 metres, "has been fiddled with" by BT to go to 300 metres and as far as 400 metres.
Basically what they are saying is a technology that was purposely designed to run at low power over a very short distance with little cross talk and radiated interference in order to provide much higher guaranteed minimum speeds (i.e. minimum of 300Mbps), will now be fiddled with to run over 400 metres with all the problems of DSLAMs having to mess with parameters, cross talk meaning your speeds only get slower as more people come online, vectoring problems, big holes in the frequencies used to avoid interference with FM radio, big chunks missing to avoid interference with ADSL and VDSL that run along side, just so a few lucky people close enough get a significant increase in speed (those within 75 metres that G.fast was designed for), and everyone else on the margins still trailing way behind with a few extra meg and a higher monthly bill.
Congratulations BT.
Regards
Phil
I fully agree, it's a shambles. Though you'll not find much support with this sentiment on these forums - there appear to be quite a few people here that defend the aluminium/copper POTS used by BT to the bitter end, regardless of how unsuitable it is in this day and age.
Fingers crossed Virgin Media continue their so far successful network expansion. Since they got bought out by Liberty Global they've gone from strength to strength.
-
Liberty Global is a mixed bag.
They have definitely moved VM forward in terms of rollout progression, before the buyout VM seemed financially crippled, which was brilliant for BT. Now they doing infill's and rolling out to new areas, this may force BT to do things they otherwise wouldnt have done. It would be great also if a large company backed someone like cityfibre, enabling them to cover all major cities, which would really make BT's execs shake in their boot's.
The bad news about LG is they still wont invest in the bad areas, which remain poor to this day, there has even been VM staff members saying its harder to get capacity upgrades approved than it was pre buyout.
-
G.Fast was given speed targets at various distances from 50m to 250m (150Mb @ 250m, 200Mb @ 200m, 500Mb @ 100m, 500Mb @ 50m with VDSL2 frequencies left alone, and allowing a Max of 1Gb @ 0m). It was never /just/ a 75m system.
G.Fast News (http://gfastnews.com/index.php/90-r/238-bt-data-g-fast-is-working) has some stats/pie charts from BT showing:
75% of the G.fast lines in BT's trial are delivering > 300Mbps downstream and 30-50Mbps upstream, Trevor Linney told the Paris G.fast Summit. The majority of lines are less than 150 meters. 17% of the lines delivered between 200 & 300 megabits; 10% 100-200 megabits; 3%, 100 megabits.
38% of lines were < 100 meters; 35% 100-150 meters; 14% 150-200 meters. 8% 200-250 meters; 5% > 250 meters.Current systems are not delivering the goal of 300 megabits 300 meters but Trevor Linney is confident that improvements already on the way will bring BT close to the plan.
They also have some guesstimates (http://gfastnews.com/index.php/90-r/237-bt-s-cost-for-10m-lines-g-fast-1-2b) of rolling out g.fast in the UK based on how many cabinets TBB estimate Openreach need to enable
-
G.fast trial customer feedback
[youtube]https://www.youtube.com/watch?v=dJYqDNnrZaE[/youtube]
Three PlayStations, two gaming PCs, smart TVs and a 4K TV? Right then... ::)
-
This is a video showing how our American friends are preparing to use G.fast too.
Bourns Video Series – G.fast Communication Solutions
Bourns G.fast line driver protection solution includes the new Bourns® TISP4G overvoltage protector, specifically-designed G.Fast transformers, and a large family of Bourns® Gas Discharge Tubes.
[youtube]https://www.youtube.com/watch?v=2raIZNodl80[/youtube]
-
This is a video showing how our American friends are preparing to use G.fast too.
I bet our American friends will have a comprehensive rollout of G.Fast DP's, versus our (in my opinion) pathetic cabinet based nonsense.
-
The most recent stuff suggests that the Americans might be more interested in FTTB variants, for MDU's that already have coax pre-installed, but no cable supplier feeding it. Then using G.Fast over the coax.
-
The estimated range and speeds are out for the G.Fast revision lets just hope it delivers. :fingers:
(https://s31.postimg.org/txq061q3v/image.jpg)
-
The estimated range and speeds are out for the G.Fast revision lets just hope it delivers. :fingers
Worth restating that the graph come from an article: http://www.electronics-eetimes.com/design-center/advanced-simulations-gfast-vectoring-co/page/0/2 - and are based on a simulation tool not working silicon so we will need to see what is actually achieved.
-
Good to see some idea of the improvements coming. A bit disappointing to have no mention of, say, the amount of spectrum needed to get these speeds, or the gauge of copper.
In this case, perhaps the interesting part is the simulator that gives you an idea of the interplay by adding multiple cabs in an area, or in figuring out where to add extra G.Fast cabs.
-
I'm pretty sure the graphs are for standard BT conditions of 0.5mm copper and FTTC present.
-
I'm not so confident.
The highest speed, with shortest copper, rarely gets even close to 1Gbps when the 2-23MHz spectrum is being excluded. If the orange line started around 800Mbps, I'd believe it.
And, as the work is done by a German institute, wouldn't we expect them to perform work one a cable model closer to home? 0.4mm seems common in Europe, though I think I've seen 0.6mm is common in Switzerland. Austria is also home to a fair amount of G.Fast interest.
There are lots of conditions that apply to any g.fast graphs, and none of them are specified here. I would be happier with a graph like this one from Swisscom in 2015, with all the attendant details:
(https://s32.postimg.org/9ntc2m8v9/Swisscom_GFast_Speeds2015.png)
I'd love it if we could get graphs that show the new results (increased bit, power and PSD) with the same level of detail. Especially if it came with an idea of the cable model:
(https://s32.postimg.org/i472aor3p/Swisscom_Cable_Models2015.png)
It is interesting to see from that model that (in Switzerland, at least) the 0.5mm "standard" copper has less attenuation than 0.6mm copper which is insulated by paper. It just goes to show that the copper alone isn't the only factor!
I *think* the reference "CAD55" cable is actually one of BT's cables - a 4-pair dropwire, with 0.5mm copper. If so, then BT's results are likely to be (currently) just below the blue dots on the Pe4D graph, but better than the Paper graph.
-
It's actually quite depressing that we're finding ourselves discussing the characteristics of in some cases decades old twisted pair still due to it being, by a mile, the main delivery mechanism of connectivity into our homes in 2016.
-
The good thing about VM pushing forward with their cable/ftth network ideas is that BT will have to do something or they will fall behind. I'll sure BT's ego can't take that :)
-
It's actually quite depressing that we're finding ourselves discussing the characteristics of in some cases decades old twisted pair still due to it being, by a mile, the main delivery mechanism of connectivity into our homes in 2016.
Thankfully I've more going on than to ever get depressed about 'twisted pair cables being the main delivery system' ........ but I take your point.
There must be plenty of other depressed folk around as to quote BT CEO (Gavin) .... "The UK is the most digitally advanced nation in the G20, but further investment is required if it’s to keep and extend that lead. That’s why we are poised to invest a further £6bn over the next three years, but we want the regulatory uncertainty to come to an end."
Thankfully, we should maintain our position with a further £6 Billion investment.
[/i]
-
The good thing about VM pushing forward with their cable/ftth network ideas is that BT will have to do something or they will fall behind. I'll sure BT's ego can't take that :)
If cable had not been rolled out in the 90s, wouldn't there be more of a business case for rolling out nationwide fibre now?
-
If cable had not been rolled out in the 90s, wouldn't there be more of a business case for rolling out nationwide fibre now?
There would probably be more of a business case as far as the country is concerned, but probably less incentive as far as BT is concerned.
I read an interesting quote earlier today, from someone describing why NTT in Japan had a high amount of fibre, and the answer came back as: the incumbent (NTT) was losing too much market share, and needed fibre to bolster themselves. Same for Verizon, and why they went for FiOS.
It seems that competition is indeed a driver, but it has to be *serious* competition with serious infrastructure. VM has probably counted as merely "semi-competent" at competition so far, but there are signs that could change, now they woken up again.
Full quote:
Q: how was that (fibre deployment to enable FTTH) achieved in Tokyo?
A: The same way that it was throughout the FiOS properties. NTT deployed the fiber in underground and aerial (whatever is appropriate).
The challenge of fibre deployment is the business case. The number 1 factor in the case is NOT construction. It is line loss. In the case of Verizon, they were under extreme pressure from Cablevision (in particular) and other MSOs. Not only were they losing on the broadband front, they were losing phone lines. FiOS was built to stem that tide and has been wildly successful from that standpoint.
Why didn't AT&T follow suit? If you look at where AT&T is (Texas, California, now the South East) there was more new lines being installed and not as much pressure on lines. So, AT&T went the U-verse route - trying to save some money. They are working to play a "lose slowly" game for residential broadband. They have ftth available for markets that they want to hotly contest, but have not done something ambitious.
NTT was getting its head handed to it by the alternate DSL providers in Japan. FTTH was a way for NTT to build a network that it didn't have to unbundle and share. So, it was a lock out the competition move. This is somewhat different than Verizon - who did have some CLEC competition but its primary competition was from MSOs.
The whole highlight here is that competition can drive FTTH deployments. Note the utter lack of wireline compeition in Europe. In spite of article trumpeting FTTH taking off in Europe on this site for the last 10 years based on pronouncements from service providers, almost nothing has happened.
The only other way to get FTTH built is through governement intervention. South Korea is an example of that. Australia should have been an example of that but has fallen flat.
Back to the business case, Construction Costs is the #2 factor...which is why Verizon started in areas with Aerial Plant.
And as a note, none of this has changed for the past 15 years.
(US-UK Glossary: MSO = cable company; CLEC is a non-incumbant telco, quasi-LLU).
The advantage to NTT was that it wouldn't have to unbundle or share fibre. Imagine if BT were offered that by Ofcom?
-
If BT was offered unbundled and unshared FTTP we would all have FTTP by next week, a bit stupid seeing as BT Openreach should make the same amount of money regardless of the reseller :-X
Maybe if things get bad enough theyll privatise it Network Rail style and subsidise ISPs like they do rail companies ::)
I just want me some G.Fast in the short term :D, I thought the original idea of a G.Fast node on every Telegraph pole was an ideal compromise vs full FTTP, everyone could get 300-500mbps and people who wanted more could pay a few £100 to get FTTPoD.