In reply to a post by Fido:

Using my own router would allow me to reset the noise margins, (at least it would have been possible with my Asus DSL-AC68U router that I sold after joining Sky); are there other ways to reset the noise margin ?

Forcing the SNR Margin on VDSL in the way that people did often on ADSL generally makes things worse on VDSL. The DLM notices this is happening and either forces the DSLAM to ignore or starts throttling the line. VDSL is not ADSL and you don't necessarily have a 6db SNR margin downlink in the way ADSL did.

The point is the openreach DLM is independent of the ISP and the DSLAM in the cabinet is trying to keep ALL the lines it is handling at the best they can do, within the problems of crosstalk. Crosstalk is a real issue on VDSL and quite different to ADSL - this is why people were hoping for Openreach to roll out vectoring as a method to counteract crosstalk.

However Openreach have abandoned the 'active NTE' for all users (the white box modem) and gone to please the public with one box solutions, which means we won't have DSLAM and modems that are managed by the infrastructure provider - so vectoring will probably never happen. Missed opportunity.

Roll on G.FAST.

In reply to a post by jchamier:

vectoring will probably never happen. Missed opportunity.

Roll on G.FAST.

It seems that Openreach have decided the costs of upgrading the VDSL2 DSLAMs for vectoring are not generally justified. They have installed vectoring in some DSLAMs to help them meet BDUK coverage targets and there might be other circumstances where have or will install vectoring (especially if we finish up with a 10 Mbit/s USO), but it seems there will be no widespread roll-out of vectoring.

Rather than spending money on VDSL2 vectoring, Openreach seem determined to focus investment on other technologies, including G.FAST. As I understand it, vectoring is a mandatory part of G.FAST.

In reply to a post by David_W:

It seems that Openreach have decided the costs of upgrading the VDSL2 DSLAMs for vectoring are not generally justified. They have installed vectoring in some DSLAMs to help them meet BDUK coverage targets and there might be other circumstances where have or will install vectoring (especially if we finish up with a 10 Mbit/s USO), but it seems there will be no widespread roll-out of vectoring.

It probably doesn't help that one of their two vendors doesn't support vectoring without a hardware swap (ECI I think). However its a disappointing position.

Rather than spending money on VDSL2 vectoring, Openreach seem determined to focus investment on other technologies, including G.FAST. As I understand it, vectoring is a mandatory part of G.FAST.

I also read that it was part of the G.FAST spec so couldn't be left out of hardware implementation. I'm hoping (being about 550m from my VDSL cabinet) that G.FAST may arrive in my area. Eventually !

In reply to a post by jchamier:

It probably doesn't help that one of their two vendors doesn't support vectoring without a hardware swap (ECI I think). However its a disappointing position.

I haven't been following the discussion on exactly what is needed to support vectoring. From what I recall, some ECI VDSL2 DSLAMs would have to be replaced to support vectoring. I believe Huawei VDSL2 DSLAMs require certain boards to be certain versions as well as the vectoring hardware to be installed, implying that many otherwise serviceable boards cannot be used in a vectoring configuration. It has been suggested online that a fully loaded Huawei VDSL2 DSLAM in the 384HD configuration cannot support vectoring, possibly because the higher density line cards used do not support vectoring, possibly because the vectoring engine and/or backplane bandwidth are insufficient to support vectoring across so many ports.

Even if it is merely a case of slotting in the vectoring engine board and making any necessary configuration changes, this does not come for free. There is almost no payback if Openreach to install VDSL2 vectoring unless vectoring helps them meet a coverage target. A small number of people might upgrade to higher speed products if vectoring makes the upgrade worthwhile, but this will bring a negligible amount of extra income. Indeed, if installing VDSL2 vectoring gave people higher speeds, that would reduce the incentive to upgrade to G.FAST when that technology became available.

My view is that whilst it is disappointing that Openreach will not be making a widespread VDSL2 vectoring roll-out, getting G.INP working on ECI DSLAMs and allowing stable lines to adopt a <6dB signal to noise ratio margin - both of which are coming over the next few months - will get the best out of the fielded VDSL2 hardware. At this point, I'd rather see investment targeted on coverage and on next generation technologies like G.FAST than on the essentially mature VDSL2 products.

Edited by deleted (Thu 02-Mar-17 23:18:07)

interesting.

do you mean jumping between stability, standard and speed or do you meaning jumping between different banded profiles?

I cannot disagree, the G.INP on my line made it incredibly stable, and removed the summer/winter variation. I'd like more upload (9173kbps uplink sync today) for my distance from cabinet (downstream attenuation 21.1), but I can't really complain for the price!

Prior to the legal court case, DLM was able to move several profiles (rate cap and error protection levels) in a single step. This was removed due to the legal action and and a single step up/step down method was implemented in its place.