I don't know about the specific answer to your question re: how to get stats 
from the ENB.  We have just been collecting those stats directly from the CPEs 
with our ACS and feeding them into Cacti.  Of course, if the connection is *so* 
craptastic that you can't probe the CPE directly during one of its fits, then 
this doesn't help.

But I did want to ask you: what subframe profile are you using?  I ask because 
this sounds suspiciously like the issue we experienced that I have mentioned 
here at least once in the past: when using subframe profile 1, we would 
encounter random terrible performance to random CPEs.  Rebooting the ENB and/or 
specific CPE would not necessarily cure the problem: the problem came and went 
as it pleased.  Sometimes the terrible performance would last for just a few 
minutes, and sometimes for hours.  It would not affect an entire sector, and it 
was happening all over the network (not just in specific geographies, which 
would seem to rule out interference from neighboring networks; also, there are 
no neighboring 3.6GHz networks).

I don't believe I mentioned this here in the past, but we did also notice 
strange see-saw spike and valley patterns on the CINR to a specific CPE when 
the issue was in effect.  So your issue sounds eerily similar.

We realized that there was a correlation between when we were advised to switch 
to subframe profile 1 and when the problems started.  So we reverted back to 
profile 2 and the problem vanished completely.  I am still a little frustrated 
about this because the sense I have is that we need more time dedicated to 
upload, which profile 1 was giving us, but we don't have the option to use that 
profile because of this issue.

We never got a satisfactory answer about why this might be the case.  We were 
just advised to stick with profile 2 if that's what worked the best for us.

-- Nathan

From: [email protected] [mailto:[email protected]] On Behalf Of 
Jeremy Austin
Sent: Friday, December 09, 2016 12:12 PM
To: [email protected]
Subject: [Telrad] ENB CINR stats?

Does anyone know how to check the equivalent of CINR and associated MCS from 
the ENB or EPC side?

We have UEs that show good CINR on the UE itself, let's say, mid-20s.

As a somewhat moving target, occasionally a UE will start dropping ICMP and 
have poor throughput measured via iperf. We've been tracking on this since 
August and have found no solution yet.

Telrad support proposed that we are seeing jittery MCS on the upstream channel. 
I'm not sure how to measure this, or see the equivalent CINR/SINR stats for a 
particular UE from the ENB's side.

Thoughts?

--
Jeremy Austin

(907) 895-2311
(907) 803-5422
[email protected]<mailto:[email protected]>

Heritage NetWorks
Whitestone Power & Communications
Vertical Broadband, LLC

Schedule a meeting: http://doodle.com/jermudgeon
_______________________________________________
Telrad mailing list
[email protected]
http://lists.wispa.org/mailman/listinfo/telrad

Reply via email to