The second phase of BTOpenreach’s VDSL Vectoring (ITU-T G.993.5) trial, which could improve the performance of BT’s existing “up to” 80Mbps capable hybrid-fibre (FTTC) superfast broadband lines by significantly reducing interference (crosstalk), has finally got underway and will run for “at least” 3 months.
ISPreview.co.uk first detailed Openreach’s plan for the second trial during May 2014 (here) and we recommend reading that for a better perspective. Suffice to say that the first trial, which began last September 2013 and continued for several months, proved to be broadly successful by enabling many more customers to achieve the top speed of 80Mbps (unofficially others reported a general boost of 10-15Mbps+ to ISPreview.co.uk).
However it’s important to stress that experiences do vary and those likely to see the most benefit are customers that have already lost a large chunk of performance due to the crosstalk problem (the more FTTC lines, the greater the problem [interference]). In other words, Vectoring is arguably more about giving end-users the performance that they should have been able to receive in the first place and less about boosting the headline rate beyond 80Mbps, although it would make an increase more feasible (e.g. 100Mbps+).
The second phase of Openreach’s Vectoring trial is set to once again focus upon the same three street cabinets [26, 41 and 42] in Barnet (London) and another three [12, 39 and 74] in Braintree (Essex), which could be expanded at a later date. Several new enhancements are also being introduced to the trial in order to help improve FTTC performance and tackle interference.
Phase 2 Vectoring Improvements
* Physical Retransmission (G.INP)
Openreach has already conducted a separate “limited trial” of G.INP with a small number of FTTC/VDSL lines (here), which is designed to tackle spikes / burst of electromagnetic interference (impulse noise) and can thus make some problematic lines more stable and less prone to errors (Sky Broadband already uses G.INP).
* Seamless-Rate-Adaptation (SRA)
This could help to keep lines stable by varying the speed more effectively. A more technical explanation is to say that SRA can be used to “reconfigure the total data rate by modifying the framing parameters and the bits and fine gains parameters“. If the noise condition improves then SRA can be used to gradually increase the data rate again, potentially taking it all the way back up to what you had before a nasty bout of sudden noise hit your line.
* Application Specific Integrated Circuit (ASIC)
The enhanced vectoring engine (ASIC) is capable of cancelling out significantly more cross-talk interferes per line, which means end-users could see an even bigger improvement in their service speeds. But it may also carry a bigger cost due to the greater need for some hardware replacement
Separately Openreach still wouldn’t be drawn on whether they eventually intended to roll-out a System-Level or Node-Level Vectoring solution, with the latter arguably being better for big cabinets with lots of FTTC lines where the system-level method might struggle.
A spokesperson for Openreach simply told ISPreview.co.uk, “We’re currently focused on ensuring that the trial of the ASIC engine is successful. Beyond that we will be considering all commercially viable options to further enhance the customer’s experience.”
Elsewhere Openreach informs that the second trial will start with the activation of 3 ASIC vector engines at the sites in Braintree and Barnet, while a further 3 will be activated at the trial sites over the coming weeks. Neil Isherwood, the man responsible for in-life aspects of Openreach’s trial, said: “The trial is an important opportunity to monitor how vectoring interacts with live network services“.
Matt Brotherton, Openreach’s Vectoring Project, told ISPreview.co.uk:
“This marks a key milestone for the project as we begin to test strategic vectoring hardware in the live network. The trial will be run in 4 phases, with the later phases looking at vectoring interworking with technologies such as physical layer retransmission and seamless rate adaption“.
Naturally consumers will be more interested in whether or not their Openreach VDSL modems will need to be replaced, which did occur for a small number of units during the first trial, but most were fine or merely needed a firmware update.
However the degree to which hardware may or may not need to be changed or upgraded, both among end-users and the street cabinets themselves (not least in terms of the ECI/Huawei equipment split), will not be certain until after Openreach have completed their trials and agreed the final way forward.
Overall it’s good to see Openreach making progress, although we’re also a little surprised at just how long-winded the trial seems to have become. We expected, or perhaps hoped, that they’d be a lot further along than they are now because the crosstalk problem is steadily becoming a lot more noticeable.