Sorry I'm a bit late on this... I haven't had a chance to look through this thread so far.
My glance at the before/after "pbParams" data suggests that:
- The attenuation hasn't changed, so it doesn't look like anything catastrophic has happened to the line
- The power levels have gone up slightly, suggesting that the signal level ought to have gone up slightly
- The SNRM levels have gone down significantly.
- The sync speeds haven't changed
Because the sync speed hasn't changed, the SNRM levels can only have gone down because noise has increased. I don't think it has anything to do with the time of day of the last sync, as you still managed to get to the full sync speed anyway. The increased power means that *something* has noticed this, and tried to compensate; the higher signal level means that noise has probably increased more than the simple difference in SNRM levels.
The questions then are:
- What is the cause of this noise?
- Does it affect the line all the time?
- Does it affect all frequencies?
The more that this noise happens 24x7, the more likely it is to be crosstalk than REIN. The more it is spread over all frequencies, the more likely it is to be crosstalk too (but a little less certainly).
Before attributing things to REIN, and starting a lengthy investigation, perhaps we should consider whether the drop can feasibly be attributed to crosstalk.
My (currently) favourite documents on Vectoring give some good graphs on the impact of crosstalk (so you can then see how good Vectoring is at fixing it). One is found here
(look at graphs on page 14 and 15 for 0.4mm cable), and the other is found here
(page 9 for 0.5mm cable).
A favourite quote in there is "In an environment where several VDSL2 systems are in the same cable binder, it is FEXT (far-end crosstalk) that severely limits the performance." Just as a reminder that crosstalk is significant... and that the interference comes more from the cabinet-end, so it doesn't have to be a close neighbour to have an effect.
How significant? The graphs suggest, for 0.5mm cable, a speed drop from 80Mbps to 55Mbps (at 3000 feet, or 1000 metres). Or for 0.4mm cable, a speed drop from 70Mbps to 52Mbps (at 1800 feet, or 600 metres.
So that kind of drop *is* possible, but it probably requires more than one crosstalk-ing line.
Probably the first questions I'd ask are:
- How far are you from the cabinet?
- How much does the speed vary over a few days? Are there sharp changes in variation, or just sudden drops & increases?
- Was the drop from 70-80 down to 52 a sudden one, or could it have been spread over weeks?