Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Black Sheep]

Really I will admit I haven't seen any photo's or the like, just what is available on our internal sites and they all seem to point towards a bolt-on to the existing FTTC DSLAM ?? Would that not make more sense for the initial phase trial/roll-out ....... not got a Scooby-Doo myself ? 

[burakkucat]

. . . a bolt-on to the existing FTTC DSLAM ?? Would that not make more sense for the initial phase trial/roll-out . . .

Indeed, that would be eminently sensible for the trial. 

Ray's starting post of this thread has a link to an ISPreview article in which the photograph clearly shows a PCP with a bolt on "G.Fast pod".

[gt94sss2]

Will this do? This comes from a Sagentia report on the access network, and shows the distribution of D-side lengths; I added the actual percentage numbers after spending time with a ruler.

I've added the graph that shows the distribution of exchange line length too (E+D-sides)

That will do nicely - it's always worth reminding some that most properties are relatively close to PCPs

Ray's starting post of this thread has a link to an ISPreview article in which the photograph clearly shows a PCP with a bolt on "G.Fast pod".

TBB also covers this at http://www.thinkbroadband.com/news/7447-update-on-g-fast-and-lr-vdsl2-trials-from-openreach.html and their text clearly says "Pods will be fitted to the side of the green PCP cabinets with the power and fibre run in from the associated VDSL2 fibre twin." with matching picture of PCP/pod.

I guess the problem with using pods with DSLAMS is the extra copper length between that and the PCP - given how sensitive g.fast is to distances.

[burakkucat]

TBB also covers this at http://www.thinkbroadband.com/news/7447-update-on-g-fast-and-lr-vdsl2-trials-from-openreach.html and their text clearly says "Pods will be fitted to the side of the green PCP cabinets with the power and fibre run in from the associated VDSL2 fibre twin." with matching picture of PCP/pod.

Thank you. I knew I had read it, and seen another photograph, elsewhere. (I must say, though, that the photograph accompanying the TTB article looks to be a "PhotoShop" special!    )

[gt94sss2]

It will be interesting to see what Openreach do with cabinets where space is at a premium - such as here

I am pretty sure it was going to be recased a few months ago based on roadworks.org but nothing happened in the end. I assume they PCPs will need 'modern' casing for this to work.

[S.Stephenson]

It's funny how these are called Pods...

We finally got what we all asked for FTTP(Fibre to the pod) 

[Black Sheep]

I guess the problem with using pods with DSLAMS is the extra copper length between that and the PCP - given how sensitive g.fast is to distances.

A very valid point ......... but I would humbly suggest the vast majority of DSLAMS are located literally within spitting distance (eugh) of the original Cab. 

[Black Sheep]

Almost (if not all) documentation is currently 'In confidence' at this time, but I took a tiny snippet outlining the scope ......... "is to develop a solution to use a 48 port G.FAST MA5818 chassis unit at a PCP side pod, or in a standalone G.FAST cabinet or in the existing NGA DSLAM cabinets"

I wouldn't get hung-up and throw a hissy over the mention of a 48 port thingy ..... this may be for trials, who knows ?? The crux of the post is to show the pod will be designed to fit all 3 scenarios. 

[tickmike]

Are these in use yet  , About a month ago when I went to Chester I was walking down this road and I said look the PCP it's got a 'Baby on the side', having some strange looks from my wife and daughter also passers by  who ignored me and carried on with there conversation, odd    .

I have just looked on street view but it's a year old photo.

The next time I'm up that way in a few weeks I will take a photo this time (wished I taken one before  ).

[S.Stephenson]

I have seen similar things but they don't have the vents so I assume that they are just a regular cabinet extension.

They could put everything in place early though and just wait for the hardware to be ready.

[gt94sss2]

In the BT Community forums, an Openreach employee - while being careful what he says because of the same rules that will apply to BS - is adamant that:

initially the G.fast node will be attached to the cabinet to serve customers within 350 mtrs. That will not cover 10m homes passed by 2020. To enable Openreach to pass 10m homes then the additional G.fast nodes must be erxtended from the cabinet to the dp thereby pushing fibre further into the network.

[S.Stephenson]

I don't see it being deployed at DPs as much as I see them intercepting cable branches.

I have a rough estimation of what I feel they will aim for.

[gt94sss2]

Yes, it does seem more logical than enabling every DP and is also more along the lines of what I would expect as well.

TBB have updated their article to add that the 'longest operating G.fast line in trials is 400m and delivering ultrafast (100 Mbps and faster) speeds.'

[S.Stephenson]

The 400m seems promising if its using current hardware as there should be a revision to the spec that allows a higher power level which should boost range.

All in all Id hope that the 160mbps should reach 400-500m when the changes are made.

[WWWombat]

I wouldn't get hung-up and throw a hissy over the mention of a 48 port thingy ..... this may be for trials, who knows ??

It is indeed a temporary limitation for the current trials.

The main cause is that, right now, chipset companies can't put together the silicon that can run 96 lines - which is BT's apparent ultimate target.

It isn't the plain "running" of a line (or 96 of them together) that is the problem. It is the work needed to get the vectoring process to successfully use data for 96 lines in parallel.

Sckipio shows this most publicly, and had this to say 3 months ago. Slide 4 shows "vectoring for 16 lines shipping in 2015; vectoring for 24 lines shipping in early 2016; vectoring for 32 lines in demo form". Huawei appear to say they'll have 96-port units available at the end of the year ... so definitely not in time for the pilots.

a solution to use a 48 port G.FAST MA5818 chassis unit

Googling suggests that the MA5818 is a chassis of the same size and appearance as the MA5616 appearing in All-in-one cabinets. It has room for 4 service cards, a power card, and control/backhaul card, and a fan card. The 5616 DSLAM can be seen in this all-in-one cabinet, as just the small cream-coloured part. An example of a 5818 can be seen here.

It looks like the 5800 series DSLAMs are aimed at the NG-PON (10G PON) backhaul market, so are likely have an increase in backplane switching speeds compared to the 5600 series. The next generation of hardware, I guess.

a) a PCP side pod, or
b) in a standalone G.FAST cabinet, or
c) in the existing NGA DSLAM cabinets

Really good to see the options being kept open.