G.INP

[#1]

Now that's more like it.

[#2]

Quote from: Stoker

Following a fault last week I had a lift&shift and DLM reset., and no longer have G.IMP enabled. I have an (unlocked) Huawei modem and am on a Huawei cabinet. G.IMP was enabled before the problem. Can PlusNet enable it through profiles or do I have to wait for BTOR? Interleave depth:       13 5 INP:                     0 0 G.INP:                   Not enabled

The same happened to me, over 60 days ago. I had G.INP them after Lift&shift + DLM Reset, no more G.INP !!!
It has just been re-enabled with G.INP ver 2, as it seems to be called.
i.e. G.INP on DS only.
It will come back of its own accord but the timing is another of the Great Openreach mysteries we have to endure.

[#3]

[#4]

Well this is odd. I saw earlier that I had some more SNRM available (yes I know it fluctuates but I read G.INP might even be able to stand as low as 3dB if BT ever thought about offering it) so I gave the Huawei a power down and am now connect at 77mbps. However a speedtest is now giving a slightly slower speed even though BTw's site says my profile is now 74.6
Hmm. Thoughts?

[#5]

@deadman2020:  It wasn't broken but you fixed it anyway and now it doesn't work so good.  Even thought of working for BT (Openreach)? 
Without G.INP I'd have suggested that you may have more errors at the higher speed reducing your throughput due to retransmission being necessary but with G.INP I thought that didn't happen (not on the main bearer anyway).  Even if the Plusnet Profile hasn't caught up I can't see how that would reduce speed.  Maybe you've just randomly hit a poorer performing gateway - try disconnecting your router (not the modem) to see if a different gateway changes things.

[#6]

Yeah I just gave it another PPP kick and am now hitting 70mbps. It could just be some congestion at various times of the day where I am seeing speed drops however. I will monitor it and see. I was hoping to see a noise margin good enough to hit the magical 80/20 (I always connect at 1999 on US though)

[#7]

[Deleted - was trying to edit a previous post and clicked the wrong thing]

[#8]

G.IMP'e last night. 
I was one of the first connected to my cab in April with max atainlable about 88Mbps and missed G.IMP v1
Over last 3 months x-talk reduced my sync to 76Mbps.
BUT now with G.IMPv2 active on my line max atainable is 81Mbps
Also ping has reduced from 16ms to 10ms.
An increase of about 5mbps and reduction in ping ......Haaaaapyyyy
ATB
SN

Max: Upstream rate = 29691 Kbps, Downstream rate = 80364 Kbps
Bearer: 0, Upstream rate = 19999 Kbps, Downstream rate = 80000 Kbps
Bearer: 1, Upstream rate = 0 Kbps, Downstream rate = 0 Kbps
Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 6.2 15.2
Attn(dB): 12.0 0.0
Pwr(dBm): 13.8 1.3
VDSL2 framing
Bearer 0
MSGc: -6 26
B: 203 237
M: 1 1
T: 0 42
R: 10 16
S: 0.0809 0.3781
L: 21153 5374
D: 8 1
I: 214 127
N: 214 254
Q: 8 0
V: 5 0
RxQueue: 78 0
TxQueue: 26 0
G.INP Framing: 18 0
G.INP lookback: 26 0
RRC bits: 0 24
Bearer 1
MSGc: 186 -6
B: 0 0
M: 2 0
T: 2 0
R: 16 0
S: 5.3333 0.0000
L: 48 0
D: 3 0
I: 32 0
N: 32 0
Q: 0 0
V: 0 0
RxQueue: 0 0
TxQueue: 0 0
G.INP Framing: 0 0
G.INP lookback: 0 0
RRC bits: 0 0
Counters
Bearer 0
OHF: 0 1155665
OHFErr: 0 43
RS: 1712395464 2312051
RSCorr: 10591 592
RSUnCorr: 0 0
Bearer 1
OHF: 7596949 0
OHFErr: 0 0
RS: 91162646 0
RSCorr: 0 0
RSUnCorr: 0 0
Retransmit Counters
rtx_tx: 155 0
rtx_c: 155 0
rtx_uc: 0 0
G.INP Counters
LEFTRS: 1 0
minEFTR: 79983 0
errFreeBits: 148871705 0
Bearer 0
HEC: 0 0
OCD: 0 0
LCD: 0 0
Total Cells: 1593248738 0
Data Cells: 48918376 0
Drop Cells: 0
Bit Errors: 0 0
Bearer 1
HEC: 0 0
OCD: 0 0
LCD: 0 0
Total Cells: 0 0
Data Cells: 0 0
Drop Cells: 0
Bit Errors: 0 0
ES: 3593 1217
SES: 31 0
UAS: 63 53
AS: 122028

[#9]

@supernova That's a nice increase I'm hoping a similar thing can happen for me, i have lost around 10Mb download and 5Mb upload thanks to what i'm told is interference and cross talk, that said i'm on a ECI cab so I have no way of knowing if anything positive is going to happen, the ECI rollout is looking likely to start happing in August so will look to see how it goes once ECI cabs start getting G.INP enabled.

[#10]

Hi Terranova,
I forgot to mention,
Previous I was on fast path / no interleaving or G.IMPv1 so the improvement is due to G.IMPv2 and not because I was interleaved / running broken G.IMPv1 thats now been fixed.
I thought I might see  at best 1 or 2mbps increase, so very happy that I'm back up to the magic 80mbs.
The SNMR is the same at 6.1db
Hope it works for you too.
ATB
SN

[#11]

Quote from: supernova1

G.IMP'e last night.  G.INP Framing: 18 0 G.INP lookback: 26 0

Notice anything?   

[#12]

Quote from: Estragon

Instead of all the hassle of interleaving, which is considerable on FTTC because of the high speeds, little if any interleaving is then applied. Instead of error correction, which is what interleaving is used for, duff packets are simply re-transmitted on Bearer 1, without interrupting the main Bearer 0.

That's what we used to think.
After investigation, it looks like bearer 1 only carries an overhead channel, with status information being passed back and forth between modems. This bearer keep old-style FEC+interleaving protection (even more aggressive than standard DLM), without retransmission.
The main end-user data stream is carried in bearer 0 - both the original blocks as well as any retransmitted ones.
In old-style DLM intervention, the FEC+interleaving settings were harsh - causing 8ms or 16ms latency delays, and using up to 20% of the bandwidth,
In new-style DLM intervention (which happens to almost all lines now (edit: well, Huawei-cabinet lines, anyway) beyond the initial DLM reset period), retransmission through G.INP is the mainstay. However, FEC+interleaving remain in use, but with much less aggressive settings: around 0.2ms delay, and around 5% of bandwidth.

[#13]

Oh 'eck!

[#14]

See the attached graphs from my connection, currently on G.INP Mk2 (not active for upstream).
Bearer 0 Interleaving depth Is 8 for DS & 1 for US.
Along with the rtx retransmissions, RSCorr/FEC is fairly active on Bearer 0 & a slight amount can be seen on Bearer 1.

[#15]

<delete>