Hello again,

I will reply to my own email because I did a mistake. Sorry for any inconvenience. I mismatched the DAHDI conf with another one that had E1 config inside.

DAHDI config is fine for T1 link but OpenBSC is still is not clear if timeslot allocation is OK or maybe someone with more experience can share why they are allocated like that.

I checked the default config in osmocom-nitb and also the one attached by Caled and it is not clear for me why we have the following arrangement compared to the E/// documentation for BSS:

I would re-write Caleb's config as it is the below one for an Unconcentrated LAPD:

oml e1 line 0 timeslot 1 sub-slot 0 -->> CF is only found in timeslot 1

trx 0 rf_locked 0 arfcn 512 nominal power 24 max_power_red 12 rsl e1 line 0 timeslot 1 sub-slot 1 -->> TRXC is multiplexed only for first TRX. The other RSL/TRXC links for other TRXes use the full timeslot. rsl e1 tei 0 timeslot 0 phys_chan_config BCCH+CCCH --> air timeslot 0 runs BCCH and CCCH hopping enabled 0 e1 line 0 timeslot 1 sub-slot 2 timeslot 1 -->> air timeslot 1 runs a SDCCH8 phys_chan_config SDCCH8 hopping enabled 0 e1 line 0 timeslot 2 sub-slot 3 timeslot 2 -->> subrate 0+1 are spare for 1st TRX in LAPD Unconcentrated mode. At least the E/// docs show the following: PCM TS2(spare/spare/TCH0/TCH1) PCM TS3(TCH2 /TCH3 /TCH4/TCH5) so I interpreted that subrate 0+1 are the spare ones.

phys_chan_config TCH/F hopping enabled 0 e1 line 0 timeslot 3 sub-slot 2 timeslot 3 phys_chan_config TCH/F hopping enabled 0 e1 line 0 timeslot 3 sub-slot 3

timeslot 4 phys_chan_config TCH/F hopping enabled 0 e1 line 0 timeslot 4 sub-slot 0 timeslot 5 phys_chan_config TCH/F hopping enabled 0 e1 line 0 timeslot 4 sub-slot 1 timeslot 6 phys_chan_config TCH/F hopping enabled 0 e1 line 0 timeslot 4 sub-slot 2 timeslot 7 phys_chan_config TCH/F hopping enabled 0 e1 line 0 timeslot 4 sub-slot 3

Regards, Razvan

On 11/29/16 8:43 PM, Labs wrote:

Hello Caleb,

On 11/27/16 9:01 AM, Caleb Pal wrote:

...

Good Evening,

With the recent commits for the OM2000 protocol, I figured it was time to dust off the RBS2308 and see what would happen. I am using the latest version of DAHDI, with a TE122P T1 card. All osmocom software was pulled yesterday and installed successfully on a machine running Debian Stable.

Maybe you want to try to switch you DAHDI config from T1 to E1 first since you use the BSC config with E1. T1 uses 24 channels and E1 uses 32 channels.

Second thing is that from what I saw on the wiki and on the Harald's blog posts the Ericsson RBS units never worked fully and there is still work to be done. Would be helpful if you have some traces on Abis from a real Ericsson BSC to see what is going on there.

Regards, R.

Hello Csaba,

On 12/1/16 11:25 PM, Sipos Csaba wrote:

Hi,

Maybe its not completely related, but my experience with Nokia Site is to use full E1 timeslots (64kbit) for RSL and OML signalling, otherwise it is not working, and this is not a Nokia limitation.

Nokia did a different implementation on Abis. I'm hoping to get a Nokia soon and play with it.

Just try using:

I can't try it because I don't have a PCM based Ericsson but an IP based one and I really don't know how to make the openbsc.cfg for that one since I couldn't find anything in the wiki for it. I was trying to help Caleb and also understand the OM2000. I worked previously on RAN part with E/// and we integrated this RBSes with real BSCes and according to my notes E/// is not using full PCM timeslots (64kbit) exactly. They have 3 types of LDAP configurations for their Abis to RBS.

- Unconcentrated LAPD For E1 they leave the full timeslot 0 free for synchronization and utilize timeslots 1-31. In this mode they use a full timeslot for signaling to TRU (TRXC) and 2 other full timeslots submultiplexed each in 4x 16kbit for voice and data. OML(CF) is not mentioned specifically but it says "TRXC-0 and CF goes together on the same timeslot". The E/// naming TRH (Transceiver Handler) says that is handling all LAPD signaling, signaling BTS to MS and O&M (CF). For T1 lines there is not synchronization timeslot used and sync is done from the overall T1 line so T1 config should start with "timeslot 0".

- Concentrated LAPD Recommended for 3 or more TRUs. Here TS0 64kbit is for synchronization in E1 links, TS1 is used as 4x16kbit signaling for the first 4 TRXes, TS2-9 submultiplexed each in 4x16kbit for traffic, TS10 again 4x signaling,etc. For T1 the same but no TS0 so config starts with TS0 for 4x signaling links for the first 4 TRXes. Since in this mode you need also configuration for CON the CF(OML) is added by CON to the same timeslot as the first 4x TRXes that means TS1.

- Multiplexed LAPD This is used for small cells and here I made the mistake in my previous mail saying is "Unconcentrated". This type of config multiplexes TRX signaling + CF(OML) + BCCH/CCCH + SDCCH8 in one full 64kbit timeslot and on the second timeslot keeps 2x 16kbit subrates spare.

So in the end I am still not sure how the config should look like to have a PCM based RBS working with OpenBSC.

And don't forget to configure the BTS Abis allocation accordingly!

Also for DAHDI, you will need to use the correct config with CCS signalling. This is an example for the system.conf for an E1 based BTS:

span=1,0,0,ccs,hdb3,crc4 bchan=1-31

Thanks for config.

Regards, Csaba

Regards, Razvan