laforge has posted comments on this change. ( 
https://gerrit.osmocom.org/c/osmo-bts/+/15918 )

Change subject: l1sap: merge MEAS IND into PRIM PH DATA / PRIM TCH
......................................................................


Patch Set 1:

(3 comments)

https://gerrit.osmocom.org/c/osmo-bts/+/15918/1/src/common/l1sap.c
File src/common/l1sap.c:

https://gerrit.osmocom.org/c/osmo-bts/+/15918/1/src/common/l1sap.c@74
PS1, Line 74: static bool tch_data_meas_present = true;
> AFAIU You don't need to do any switching, simply have an API bool 
> bts_model_tch_data_meas_present()  […]
we do already have the BTS_FEAT_* constants and the gsm_bts_set_feature() API.  
That would be the most logical way for me to deal with this.

Adding a single call to gsm_bts_set_feature() in all existing backends is also 
certainly not requiring a full re-test with all hardware.


https://gerrit.osmocom.org/c/osmo-bts/+/15918/1/src/common/scheduler.c 
File src/common/scheduler.c:

https://gerrit.osmocom.org/c/osmo-bts/+/15918/1/src/common/scheduler.c@760
PS1, Line 760: l1sap->u.tch.
> I am also a bit confused about the is_sub struct member. […]
I believe I explained this elsewhere here in gerrit or in redmine.  is_sub must 
at the very least be set for all SACCH  frames, and is set on other frames 
depending on the codec that's being used.  It controls which frames count into 
RXLEV/QUAL "SUB".


https://gerrit.osmocom.org/c/osmo-bts/+/15918/1/src/osmo-bts-trx/scheduler_trx.c
File src/osmo-bts-trx/scheduler_trx.c:

https://gerrit.osmocom.org/c/osmo-bts/+/15918/1/src/osmo-bts-trx/scheduler_trx.c@a205
PS1, Line 205: /* FIXME: use actual values for BER etc */
> The problem here is that we can not use actual values for BER. […]
the best would be if those values simply wouldn't count.

Let's say you normally expect 23 values every measurement interval.  Then you 
compute the average by dividing by 23.  But if only 20 valid measurements were 
received and 3 were in bad frames, then the average should simply be computed 
by dividing by 20, and not by 23.  At this point you don't need to "invent" any 
values.



--
To view, visit https://gerrit.osmocom.org/c/osmo-bts/+/15918
To unsubscribe, or for help writing mail filters, visit 
https://gerrit.osmocom.org/settings

Gerrit-Project: osmo-bts
Gerrit-Branch: master
Gerrit-Change-Id: I710d0b7cf193afa8515807836ee69b8b7db84a84
Gerrit-Change-Number: 15918
Gerrit-PatchSet: 1
Gerrit-Owner: dexter <[email protected]>
Gerrit-Reviewer: Jenkins Builder
Gerrit-Reviewer: dexter <[email protected]>
Gerrit-Reviewer: fixeria <[email protected]>
Gerrit-Reviewer: pespin <[email protected]>
Gerrit-CC: laforge <[email protected]>
Gerrit-Comment-Date: Wed, 18 Dec 2019 12:12:21 +0000
Gerrit-HasComments: Yes
Gerrit-Has-Labels: No
Comment-In-Reply-To: dexter <[email protected]>
Comment-In-Reply-To: pespin <[email protected]>
Comment-In-Reply-To: fixeria <[email protected]>
Gerrit-MessageType: comment

Reply via email to