Hi Sylvain, hi list!
I'm experimenting with burst_ind and TCHs right now and ran into some problem I couldn't solve yet.
After receiving an Assignment Command for a hopping TCH/F I call l1ctl_tx_dm_est_req_h1() with all necessary parameters and tch_mode GSM48_CMODE_SPEECH_V1 or _EFR. After that I do get burst indications containing the received bits on up- and downlink for the active arfcn on each consecutive frame number.
BUT the rx level measurements are most of the time very low and sporadic higher, surely not from that nearby bts and the very close cellphone. It looks like the layer1 doesn't "hit" the right timeslot on the right arfcn at the right time. There are some possible sources of error leading to that, like hopping parameters, channel number and MA list. But I checked these and I took all of them directly from the ASS CMD, the MA as word list in ascending order, like in layer23 IMM ASS handling. The specific AC doesn't have any specialties like Starting Time or "before time" parameters.
So my question is if there is some obvious pitfall I'm missing and are there any suggestions how to debug that?
Regards, Mad
After receiving an Assignment Command for a hopping TCH/FI call l1ctl_tx_dm_est_req_h1() with all necessary parameters
IIRC, the L1 in burst_ind mode doesn't support changing channel assignement without first resyncing to the C0.
So you can not do BCCH -> SDCCH -> TCH you have to do BCCH -> SDCCH -> BCCH (wait for sync) -> TCH
It's a limitation of the current public code in burst_ind mode.
and tch_mode GSM48_CMODE_SPEECH_V1 or _EFR.
tch_mode is useless for burst_ind
Cheers,
Sylvain
On Tue, 15 Mar 2011 18:48:48 +0000, 246tnt@gmail.com wrote:
IIRC, the L1 in burst_ind mode doesnt support changing channel assignement without first resyncing to the C0.
So you can not do BCCH -> SDCCH -> TCH you have to do BCCH -> SDCCH -> BCCH (wait for sync) -> TCH
Ah ok, thank you, now I see frames... :-)
What does the current master layer1 do to be able to change from SDCCH to TCH without loosing sync? I didn't found it resyncing to SCH/FCCH somewhere, but perhaps I overlooked it.
BTW, after some time on a TCH it seem to loose the channel, resulting in frame loss and even complete break of reception. Am I guessing right that the missing feed of dsp measurements into afc and setting of RF gain - like it's done in prim_tch - is the cause?
Regards, Mad
i am looking also for the sync. there are so many funnctions in gsm48_rr.c so i think i have to add more than a few lines?
-- View this message in context: http://baseband-devel.722152.n3.nabble.com/burst-ind-and-TCH-tp2683277p34117... Sent from the baseband-devel mailing list archive at Nabble.com.
@27c3 the sync is done with gsm322_sync_to_cell so have a look to gsm322_sync_to_cell() @ gsm322.c perhaps someone is better in c than me and adds this function to the right place
-- View this message in context: http://baseband-devel.722152.n3.nabble.com/burst-ind-and-TCH-tp2683277p34125... Sent from the baseband-devel mailing list archive at Nabble.com.
Hi, mad
I met the same problem as you, are you settle the problem and capute the TCH correctly now?
-- View this message in context: http://baseband-devel.722152.n3.nabble.com/burst-ind-and-TCH-tp2683277p40251... Sent from the baseband-devel mailing list archive at Nabble.com.
baseband-devel@lists.osmocom.org