Print

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

[pedro492]

Welcome pedro.

Hi, kitz and all.

I find it completely unbelievable that BT group have a weekly revenue of only £2.5m on FTTC products.

It is difficult to believe. BT boasted in its latest (Q2) results that it now has 3.4 million FTTC subscriber lines. [1]     Implying revenues of just 73½p a week per FTTC subscriber!  Though that was ASSIA's estimate, and not BT's.  (see para.5 of [2] )

So maybe it was a ruse; a red herring from ASSIA's counsel.  Leave the judge questioning the 10% royalties that ASSIA claims it is due, while the gross under-estimate of BT's overall revenues from FTTC passed completely unchallenged.  At a later date, ASSIA can then point to accurate, audited FTTC revenues for BT and demand much larger damages.

Even more concerning perhaps (for BT) is the Court of Appeal ruling that its DLM system also infringes that earlier '495 patent from 2004. Relating perhaps to the DLM system used in 20CN and 21CN; i.e. the DLM system/s in BT's ADSL offerings.   If that allegation is proved correct, then it could perhaps result in a much larger damages award to ASSIA, since it would cover an infringing period of nearly a decade; involving many more DSL lines, and thus far greater revenues for BT.

From briefly reading those judgments, it was strange to find BT arguing that ASSIA had failed to reverse-engineer its DLM System, and it therefore couldn't prove that it has used an offending Profile State Transition Matrix in its implementation.  (see para.28 in [3] )

Why didn't the Court just order BT to open up its DLM System source code for independent auditing? To confirm, one way or the other, whether it has used a PSTM?   Isn't that what normally happens?  Independent code auditing?

That said, the trial judge (Mr Justice Birss) was refreshingly adept at understanding all the relevant concepts of DSL, DSM, DLM and so on.   It doesn't seem that long ago when someone had to quietly explain to a judge what the internet is! 

[1] http://www.btplc.com/Sharesandperformance/Quarterlyresults/Quarterlyresults.htm
[2] http://www.bailii.org/ew/cases/EWCA/Civ/2014/1513.html
[3] http://www.bailii.org/ew/cases/EWCA/Civ/2014/1462.html

[NewtronStar]

Hang on...are you saying BT DLM is not turned off yet? If the DLM still active then BT is lying to ASSIA in court!

The DLM cannot be be turned off thoughout the UK without BT having a secondary DLM in action if the headline is correct BT will do this in a gradual phaze over many months.

[Ixel]

Someone has confirmed that DLM is still active on FTTC connections, having had their speed banding apparently reduced by DLM yesterday.

See http://forums.overclockers.co.uk/showpost.php?p=27267941&postcount=585 over at OcUK.

[NewtronStar]

Someone has confirmed that DLM is still active on FTTC connections, having had their speed banding apparently reduced by DLM yesterday.

See http://forums.overclockers.co.uk/showpost.php?p=27267941&postcount=585 over at OcUK.

if you look back to posts by BS and Kitz. BT is slowly changing the DLM wether this is a result of this case or not BT has a plan B and as an EU you won't notice any transition from one DLM to another so the world still goes on as normal.

[pedro492]

as an EU you won't notice any transition from one DLM to another so the world still goes on as normal.

How reassuring!  No biggy, of course, as BT's share price was up at close, and that's all that really matters!
And in PR terms, how neat that this Bad News From The Courts has gotten buried beneath the excitement over BT's plans to buy-out O2.
Very Fortuitous Timing, if nothing else!

[NewtronStar]

And in PR terms, how neat that this Bad News From The Courts has gotten buried beneath the excitement over BT's plans to buy-out O2.
Very Fortuitous Timing, if nothing else!

it looks like a clear case of propaganda to me, the question is who set it up 

[kitz]

Im trying to read through the court case notes provided by pedro,  Ive still got one heck of a lot of reading to do.   Pedro have you had chance to read it all?

From what I can gather so far, much of the argument seems to be over the use of 'historic data' in deciding whether to reprofile the line.  ASSIA call this PSTM and it seems to in a way equate to BTs ILQ.  Ive mentioned ILQ in the other thread a few times.  Its what replaced blip logic and I think Ive mentioned a few times how the new method by using past records it doubles time before any further changes are made... and its why some lines take longer than others to recover from say Interleaving.

TBH Ive had to stop reading now  as I have to be up early tomorrow, but the argument over historic and the fact for FTTC data is stored on Yukon before going to Rambo seemed a bit petty [111].. but hey ho (Yukon is only used by the fttc cabs)

The point was that on one view it was the Yukon not the RAMBo which collected the data and so, by the time the RAMBo itself collected the data, it was indeed not current.

Ive still got loads of reading to do, but that would have to be left to someone with on hell of a lot more time than me. 
But my conclusion so far is that it appears ASSIA are not happy with the way BT's FTTC system stores historic data to decide when a line should be re-profiled which is similar to ASSIA's  PSTM

I dont [think] they are disputing immediate changes that the DLM makes... but just the ILQ system of  learn from past mistakes /doubler part of the system.

It all seems it may be a bit pie in the sky anyhow, because BT claim to have already changed their system so that it no longer impedes on copyright... and its now a waiting game until it goes back to court again to see if its adjudicated that the new system is OK and free from breach of copyright.   

[Chrysalis]

It reduces packet loss and retransmits so less bandwidth is wasted resending packets.

in theory.

we know it doesnt work perfect.

I expect there is many cases where DLM over reacts to a one off issue which in effect reduces performance.

[Chrysalis]

I'd imagine that might be buried in HERE somewhere:

I think you may be right in that it is 'something' to do with the vdsl system.

For those having a pop at ASSIA

I dont see that anyone has     (so far)

My personal opinion is that BT will come back to the table when they are forced to rollout vectoring - and again this is about how you MANAGE vectoring/profiling/etc, not how you implement it

The system profiles will have to change anyhow to account for vectoring.  Bet BT are busy beavers atm.

I wonder if this will motivate them into rolling out things like vectoring, who knows maybe it will help .

No it won't - BT will be the last of the Euro telecos to implement it.

Interesting regarding vectoring I Was wondering if BT were planning to merge vectoring into their DLM system, as I have read comments in places that the FTTC DLM isnt just about error control but also BT use DLM to manage PSD masks aka crosstalk control.

and yeah, it wouldnt surprise me if BT were the last EU telco to enable vectoring.  They very tip toe'y on new tech.  I think this will delay vectoring if anything.

[Black Sheep]

Vectoring is rolling out as we speak, and the DLM is/was in the process of being 'looked at' to accommodate the new technology. This is what I am reading within the confidential docs that are visible to myself.

However, there are mountains of docs surrounding DLM and its architecture. It is the most complicated thing I've ever seen. I posted elsewhere (Don't know where though) about one snippet that I came across, which states the reporting system sends up to 40 (approx. ?) events to DLM ....... some every 15mins, some every 6hrs. I seem to also remember power per-tone, SNR per-tone was also in that list of events ?

The revelation about DLM being switched off, I have no idea about ?? I can say that I would have expected the OR DSL engineering community to have been informed of such practices, if it is to be implemented ??

NB .... I won't be posting up confidential docs, so my only reference is my word I'm afraid. 

[Ixel]

Vectoring is rolling out as we speak, and the DLM is/was in the process of being 'looked at' to accommodate the new technology. This is what I am reading within the confidential docs that are visible to myself.

However, there are mountains of docs surrounding DLM and its architecture. It is the most complicated thing I've ever seen. I posted elsewhere (Don't know where though) about one snippet that I came across, which states the reporting system sends up to 40 (approx. ?) events to DLM ....... some every 15mins, some every 6hrs. I seem to also remember power per-tone, SNR per-tone was also in that list of events ?

The revelation about DLM being switched off, I have no idea about ?? I can say that I would have expected the OR DSL engineering community to have been informed of such practices, if it is to be implemented ??

NB .... I won't be posting up confidential docs, so my only reference is my word I'm afraid.

Interesting. Thanks for the insight, good to hear vectoring is rolling out - though depending on how fast that rollout is happening and hopefully includes both Huawei and ECI cabinets (last the public heard was that it was in a second trial I believe).

[adslmax]

NB .... I won't be posting up confidential docs, so my only reference is my word I'm afraid.

sound selfish not to share with us?

[roseway]

I assume you know what the word 'confidential' means!

[tommy45]

Vectoring , will this have a negative impact on latency in that vectoring will increase it ? If so they can keep it off my connection i ain't interested  in extra speed 74-75mbps is enough for now,

[Chrysalis]

Reading this article makes me think if it is off its going to be very temporary, probably early December a modified version of DLM will be reinstated.

http://www.ispreview.co.uk/index.php/2014/11/uk-court-appeal-rules-bt-fttc-broadband-infringes-assia-patents.html

Tommy vectoring is noise cancellation technology, meaning it will be less likely you will get interleaved due to less errors caused by crosstalk, if that makes sense.