Will Sonic upgrade DSLAMs to VDSL2? (if not already done?)

[droz037]

http://www.networkworld.com/news/2013/0 ... 68919.html

[dane]

Yes, we do have plans to deploy VDSL2 as soon as our equipment vendor begins to deliver the interface in volume. We have sample VDSL2 interfaces in test now in our network.

That being said, it's unclear if vectoring will be viable for a central-office delivered product. From a remote cabinet, where 100% of the end-users could be connected to the same equipment, vectoring is viable. But in a mixed environment, with other DSL, T1s, etc, it's unclear if vectoring will provide any gains. That said, vendors are continuing to evolve the implementations, and I saw a presentation from Alcatel late last year that claimed vectoring gains in the presence of non-vectored signals (legacy DSL, etc), so it may get there.

[droz037]

Dane,


Thanks for your reply. Greatly appreciated that you're active on the forums.

I live in The Sunset in SF and am anxiously waiting for Sonic's fiber. I'm two blocks from the CO at 19th & Kirkham and AT&T doesn't have U-Verse on my street....


Leo

[torstenb]

Yes, we do have plans to deploy VDSL2 as soon as our equipment vendor begins to deliver the interface in volume. We have sample VDSL2 interfaces in test now in our network.

That being said, it's unclear if vectoring will be viable for a central-office delivered product. From a remote cabinet, where 100% of the end-users could be connected to the same equipment, vectoring is viable. But in a mixed environment, with other DSL, T1s, etc, it's unclear if vectoring will provide any gains. That said, vendors are continuing to evolve the implementations, and I saw a presentation from Alcatel late last year that claimed vectoring gains in the presence of non-vectored signals (legacy DSL, etc), so it may get there.

I read somewhere that one provider needs to control all lines in order to be able to perform the calculations necessary that are need to avoid the interference between lines. The other restriction seems to be that it only makes a difference below 2,000ft. Anybody further away wouldn't be able to benefit.

[dane]

The control of all signals is the issue I was noting - obviously from a central office this is impossible. But some vendors are claiming vectoring gains in the presence of "foreign" signals, but I believe that's still a mostly-vectored environment. The gains can't be substantial in a mostly mixed binder.

VDSL2 only has gains versus ADSL2+ at less than 5,000ft. Does vectoring further reduce this?

[torstenb]

VDSL2 only has gains versus ADSL2+ at less than 5,000ft. Does vectoring further reduce this?

I briefly followed the debate on Vectoring in Germany and they were stating that it can double the VDSL speeds to 100mbit at distances of max. 400 meters (1300 ft) and with g.fast to 200 mbit at distances of max. 200 meters (650 ft). I thought I somewhere read that it offers no benefits at longer distances, but I can't remember where I read it and maybe wrong.

[dane]

Here is some info - looks like short distance:
http://www2.alcatel-lucent.com/techzine ... s-promise/

We see VDSL2 itself providing gains out to 5,000ft, and clearly bonding can double speed of any of the technologies.

[droz037]

I'll volunteer as a guinea pig for testing. I'm two blocks from the CO and would love a faster connection

[mruhstaller]

Any update on VDSL2? I'm also within 5k of the CO and would love to know if this is a possibility for me soon.

[dane]

We are still doing the lab work. There are installation, filtering and CPE issues, plus provisioning and trouble tools to update if it all works out.