ADSL2+ frustrations

[Saturday]

The confusion possibly happening here is that there are several different naming conventions for ADSL (e.g. ANSI and ITU), several different versions of ADSL, further subdivisions known as annexes, further subdivisions known as releases and several different physical implementations possible. When faced with this, routers can only make a best guess of what ADSL flavour they are connected to and typically use yet another naming convention (e.g. G.DMT, ADSL2 etc).

This all gives rise to problems such as, for example, the Sky router reporting BT's ADSL Max (Sky Connect) as G.DMT. It also reports one of Sky's LLU implementations as G.DMT. Not entirely incorrect but not correct either. In simple terms BT's implementation uses an early release (annex) and Sky uses one of the latest (annex C or D if I recall correctly).

BT's implementation implementation of G.DMT is not the same annex or implementation as Sky's.

Sky's G.DMT is actually quite close in performance/technical characteristics to ADSL2 (or more correctly ITU G.992.3/4).

[James67]

Sky's G.DMT is actually quite close in performance/technical characteristics to ADSL2 (or more correctly ITU G.992.3/4).

That's probably true for lines which are capable of no more than 8Mbit/sec, but if a line is capable of more than 8Mbit/sec over the G.dmt frequency spectrum (i.e., up to 1.1MHz) then there probably is usable bandwidth in the frequency ranges up to the ADSL2+ limit of 2.2MHz.

[Saturday]

Well that's possible but (without trawling through the documentation), if I recall correctly, I think the improvements come from "borrowing" techniques like packet aggregation and frame bursting from ADSL2/2+ rather than anything to do with making better use of the frequency bandwidth.