Print

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

[WWWombat]

I've gleaned a graph from BT that purports to show the change to error rates with the introduction of G.INP.

The graph came from a different presentation, so it is blurry, and doesn't come with proper labels. But it gives an impression.

I *suspect* that the horizontal axis show the number of ES's per 24 hours. If true, then my reading is that, pre-G.INP, 90% of lines were achieving less than 500 ES's per day. Post-G.INP, 90% of lines are achieving 20 ES's per day or less.

[kitz]

Thanks for that. As you say unfortunately its a bit blurry and I can't make out any figures.
Pressumably the Y axis will be the % of lines so in that case I would tend to agree that the horizontal axis is showing the amount of errors.

For some reason that blurry graph looks familiar but I cant recall why.  Perhaps I disregarded it whilst reading something else.  The figures sound familar too, but Ive drawn a blank from where I may have seen it.

The only graphs I can find are the alcatel ones

Interesting whilst doing a search I stumbled accross the ASSIA document again and note that their system doesnt alway recommend the use of retransmission and that for very short noise bursts then FEC and Interleaving would be the preferred choice.

Retransmission may serve as the preferred
approach for mitigating impulse noise on certain
lines. In other cases, it may cause adverse
effects and should be replaced by traditional error
correction methods (see Figure 1).  Retransmission
is a powerful new tool which needs to be managed
dynamically to extract maximum performance and
stability out of a DSL network.

link

With this in mind it just goes to show I guess there is no one solution fits all.

[GigabitEthernet]

My line will probably obscure those figures quite considerably when G.INP is enabled