- OpenBSC - lists.osmocom.org

nanoBTS for sale
by Martin Mayoral Zuloaga 29 Jan '11

29 Jan '11

2 nanobts for sale $ 1,500 usd each. Both $ 2,700

1 0

please verify my multidrop-config
by Thomas Ansorg 24 Jan '11

24 Jan '11

hello list, i'd like you to go with me thru my setup of 2 bs11 and verify it. it looks like this: E1 E1 PC------BTS0------BTS1 let's begin with the bs11_config outputs bts0: LMT LOGON: ACK PHASE: 3 Normal MBCCU0: Load MBCCU1: No Load Abis-link: Restoring BS11 ATTRIBUTES: BS-11 ESN PCB Serial Number: 001104 BS-11 ESN Hardware Code Number: 135-2044/03.07 BS-11 ESN Firmware Code Number: 135-2044/03.07 PLL Set Value=1006, Work Value=1038 SITE MANAGER ATTRIBUTES: E1 Channel: Port=0 Timeslot=1 (Full Slot) TEI: 25 BS11 Line Interface ATTRIBUTES: PLL Mode: Standalone BS11 CCLK ATTRIBUTES: CCLK Accuracy: High (1) CCLK Type=0 BS11 Power Amplifier 0 ATTRIBUTES: TRX Power: 30mW (GSM) BPORT0 ATTRIBUTES: Line Configuration: Multi-Drop (1) BPORT1 ATTRIBUTES: Line Configuration: Multi-Drop (1) LMT LOGOFF: ACK and that's bts1: LMT LOGON: ACK PHASE: 3 Normal MBCCU0: Load MBCCU1: No Load Abis-link: Restoring BS11 ATTRIBUTES: BS-11 ESN PCB Serial Number: 001112 BS-11 ESN Hardware Code Number: 135-2044/03.03 BS-11 ESN Firmware Code Number: 135-2044/03.03 PLL Set Value=1073, Work Value=1073 SITE MANAGER ATTRIBUTES: E1 Channel: Port=1 Timeslot=17 (Full Slot) TEI: 25 BS11 Line Interface ATTRIBUTES: PLL Mode: Standalone BS11 CCLK ATTRIBUTES: CCLK Accuracy: High (1) CCLK Type=0 BS11 Power Amplifier 0 ATTRIBUTES: TRX Power: 30mW (GSM) BPORT0 ATTRIBUTES: Line Configuration: Star (0) BPORT1 GET ATTR NACK <--- do i have to worry about his? LMT LOGOFF: ACK here we go with openbsc.cfg. i use only trx0 of each bts. don't ask why i did something like this or like that, i just adapted the openbsc.cfg-file from the packet to my timeslots configured ! ! OpenBSC configuration saved from vty ! ! password foo ! line vty no login ! network network country code 262 mobile network code 10 short name Y-Phone long name Yodaphone handover 1 bts 0 type bs11 band GSM900 cell_identity 1 location_area_code 1 training_sequence_code 7 base_station_id_code 63 oml e1 line 0 timeslot 1 sub-slot full oml e1 tei 25 <-- i use tei 25 for the oml of bts1, too. correct? trx 0 arfcn 121 max_power_red 0 rsl e1 line 0 timeslot 1 sub-slot full <-- the same e1 timeslot for oml is used here for the ccch. rsl e1 tei 1 | correct? timeslot 0 | i'd say weird phys_chan_config CCCH+SDCCH4 | e1 line 0 timeslot 1 sub-slot full <-- timeslot 1 phys_chan_config TCH/F e1 line 0 timeslot 2 sub-slot 1 timeslot 2 phys_chan_config TCH/F e1 line 0 timeslot 2 sub-slot 2 timeslot 3 phys_chan_config TCH/F e1 line 0 timeslot 2 sub-slot 3 timeslot 4 phys_chan_config TCH/F e1 line 0 timeslot 3 sub-slot 0 timeslot 5 phys_chan_config TCH/F e1 line 0 timeslot 3 sub-slot 1 timeslot 6 phys_chan_config TCH/F e1 line 0 timeslot 3 sub-slot 2 timeslot 7 phys_chan_config TCH/F e1 line 0 timeslot 3 sub-slot 3 bts 1 type bs11 band GSM900 location_area_code 2 training_sequence_code 7 base_station_id_code 63 oml e1 line 0 timeslot 17 sub-slot full <-- harald told me line has to be 0, but i dont understand why. i configured 2 fragments on the misdn driver, with d-channel on ts 1 and 17, the rest b-channels. each fragment is a device,right? (0 and 1). Each device corresponds to a port in bs11_config and a line here, right ? oml e1 tei 25 <-- here again: all oml e1 teis identical? trx 0 arfcn 122 max_power_red 0 rsl e1 line 0 timeslot 17 sub-slot full rsl e1 tei 2 timeslot 0 phys_chan_config CCCH+SDCCH4 <--here we go with line(=port=id) 1, since fragment 2 starts ad device id 1 on e1 ts 17 as d-channel e1 line 1 timeslot 17 sub-slot full timeslot 1 phys_chan_config SDCCH8 e1 line 1 timeslot 18 sub-slot 1 timeslot 2 phys_chan_config TCH/F e1 line 1 timeslot 18 sub-slot 2 timeslot 3 phys_chan_config TCH/F e1 line 1 timeslot 18 sub-slot 3 timeslot 4 phys_chan_config TCH/F e1 line 1 timeslot 19 sub-slot 0 timeslot 5 phys_chan_config TCH/F e1 line 1 timeslot 19 sub-slot 1 timeslot 6 phys_chan_config TCH/F e1 line 1 timeslot 19 sub-slot 2 timeslot 7 phys_chan_config TCH/F e1 line 1 timeslot 19 sub-slot 3 and for the final: the output of bsc_hack during start: 2 devices found id: 0 Dprotocols: 00000018 Bprotocols: 0000000e protocol: 4 nrbchan: 14 name: hfc-e1.1-1 activate bchan activate bchan DB: Database initialized. DB: Database prepared. since line 1 isnt used for bts1 oml, only device id 0 with d-channel on ts 1 is activated. but im quite sure this is a mistake, i need 2 devices for 2 bts. right? this config dont work at all. bts1 not recognised, bts0 is recognised but cant handle calls. the handys are callable and ring, but no data is transmitted. please give me as much feedback as possible mni tnx T. -- Wer Rechtschreibfehler findet, darf sie behalten!

1 0

how to create trx0?
by Thomas Ansorg 21 Jan '11

21 Jan '11

hello, my 2nd bs11 has no load on trx0. may someone give me a bit more detailed information than on openbsc.org how to create trx0? tnx in advance T. -- Wer Rechtschreibfehler findet, darf sie behalten!

2 3

Re: State of the HSL Femtocell support
by Dakota C 19 Jan '11

19 Jan '11

>You would have to talk to the manufacturer and its distributors about >this. I have no commercial or other relationship with the manufacturer. I myself tried this route as I'm a student looking for an inexpensive way to work with OpenBSC, but... 1: they require a mutual shared NDA before they'll even talk prices; 2: they claim it REQUIRES their proprietary BSC software and thus you must purchase a femtocell developer kit including their A-bis specs and a bunch of stuff you don't need, inflating the price; and 3: they only offer their NDA to companies -- individuals and students need not apply. 4: ... but they DO ship internationally -- GSM1800 only. But looking at the /only/ price they mention - $7200 US for two femtocells and all the software -- why bother? You're better off buying two, maybe three NanoBTS for that price. And I also seem to question my own point #2 -- they first tell me "If you do want to use the HSL Femtocell with an alternative to the HSL BSC..." and then in the latest mail "... our femtocell is designed to be used with our own BSC" making it seem as if you're not supposed to be taking it elsewhere. Le sigh. -DC

1 0

cellmgr_ng moved to GNU AGPLv3
by Holger Hans Peter Freyther 16 Jan '11

16 Jan '11

Hi all, I have moved the cellmgr_ng code to GNU AGPLv3 as well. I am going to land M2UA code, a simple STP tomorrow and also the linkset cleanups I have been doing.

1 0

Perl SIGTRAN stack
by Duncan Salerno 16 Jan '11

16 Jan '11

Hi, Found a perl (!) SIGTRAN stack on the net, thought someone here might be interested. Includes a very basic HLR emulator and various other bits. Looks fairly easy to extend to send/respond to various GSM MAP messages, could be useful for testing. http://mtve.livejournal.com/13725.html http://frox25.no-ip.org/~mtve/wiki/SigTranSol.html Enjoy! Duncan

2 1

cellmgr_ng on the way to a STP and MGW
by Holger Hans Peter Freyther 15 Jan '11

15 Jan '11

Hi all, the cellmgr_ng code has gained M2UA handling and is on the way that it could play the classic role of a Media Gateway (MGW) as of RFC3331, or a plain Signalling Transfer Point. Is someone interested in helping to finish that up and using it somewhere? regards holger

1 0

[ADM] openbsc mailinglist address change
by Harald Welte 15 Jan '11

15 Jan '11

Hi! The 'preferred' address for the openbsc mailinglist has changed to openbsc(a)lists.osmocom.org and http://lists.osmocom.org/ However, since the latter is just a virtual instance on the same list server, the old addresses will continue to work. Please note, though, the 'List-Id' header will change the doamin name, so you may have to update your message filter scripts. Sorry for any inconvenience, but I'm just making sure all the GSM network related projects are on osmocom.org (like the recent git repository move). Regards, Harald -- - Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)

1 0

BTS config problem
by Thomas Ansorg 15 Jan '11

15 Jan '11

hello im still trying to make 2 bts work in multidrop mode. bts0, the one on the pc, tells that: LMT LOGON: ACK PHASE: 3 Normal MBCCU0: Load MBCCU1: No Load Abis-link: Restoring BS11 ATTRIBUTES: BS-11 ESN PCB Serial Number: 001104 BS-11 ESN Hardware Code Number: 135-2044/03.07 BS-11 ESN Firmware Code Number: 135-2044/03.07 PLL Set Value=1006, Work Value=1038 SITE MANAGER ATTRIBUTES: E1 Channel: Port=0 Timeslot=1 (Full Slot) TEI: 25 BS11 Line Interface ATTRIBUTES: PLL Mode: Standalone BS11 CCLK ATTRIBUTES: CCLK Accuracy: High (1) CCLK Type=0 BS11 Power Amplifier 0 ATTRIBUTES: TRX Power: 30mW (GSM) BPORT0 ATTRIBUTES: Line Configuration: Multi-Drop (1) BPORT1 ATTRIBUTES: Line Configuration: Multi-Drop (1) LMT LOGOFF: ACK this bs11 works pretty well as single config. but bts1 tells that: LMT LOGON: ACK PHASE: 3 Normal MBCCU0: No Load MBCCU1: Load Abis-link: Restoring BS11 ATTRIBUTES: BS-11 ESN PCB Serial Number: 001111 BS-11 ESN Hardware Code Number: 135-2044/03.03 BS-11 ESN Firmware Code Number: 135-2044/03.03 PLL Set Value=1042, Work Value=1221 SITE MANAGER ATTRIBUTES: E1 Channel: Port=1 Timeslot=17 (Full Slot) TEI: 25 BS11 Line Interface ATTRIBUTES: PLL Mode: Standalone BS11 CCLK ATTRIBUTES: CCLK Accuracy: Medium (0) CCLK Type=0 BS11 Power Amplifier 0 ATTRIBUTES: TRX Power: 2W (GSM) BPORT0 ATTRIBUTES: Line Configuration: Star (0) BPORT1 GET ATTR NACK LMT LOGOFF: ACK as you can see, i configured the 2nd bs11 with ts 17 on port 1, according the bmask and dmask settings(sorry for that yesterday, but sometimes there are so many trees you dont see the forest - german saying), and bsc_hack starts now without problems, but bootstrapping is only on bts0, bts1 isnt recognised. when i connect bts1 to the pc and run it in single mode(with line 1 timeslot 17 settings in openbsc.cfg), it is recogised with bootstrapping, but dont function. it isnt seen by any handy. i put 2 attachments in the mail, hope they arent dropped could the different pll and work values be the reason? is the BPORT1 GET ATTR NACK from bts1 a possible reason? what about the strange messages seen in the bts1_err.jpg? T. -- Wer Rechtschreibfehler findet, darf sie behalten!

2 1

OsmoSGSN status report
by Harald Welte 15 Jan '11

15 Jan '11

Hi all! Since more people are currently looking into using OsmoSGSN, let me write down some technical notes on its current status: NS (08.16) implementation (BSS and SGSN side) * complete + stable * supports NS/UDP/IP and NS/FR/GRE/IP stacking BSSGP (08.18) implementation * no support for per-BSS/per-MS flow control yet, this is WIP * otherwise complete and stable LLC (04.64) implementation, network side * No support for Async Balanced Mode (ABM), as to the best of my knowledge, it is not used if IP is transported over GPRS * Encryption stub functionality using external plugins for actual GEA encryption. GEA3 reference code is published, but we cannot redistribute it * No full support for the XID mechanism yet SNDCP (04.65) implementation * Fragmentation/defragmentation implenented but not widely tested * No full support for the XID mechanism yet * No support for any of the header compression formats * is optional anyway, but helps performance * No support for V.42bis payload compression * is optional anyway, but helps performance GPRS Mobility Management (04.08) implementation * Supports GPRS ATTACH, RA UPDATE * Known bugs with inter-cell hand-over * No access to HLR yet, i.e. all MS are permitted to enter * No persistent storage of MM state, i.e. all state is lost if process crashes or restart * We don't yet work around a fairly common MS bug which makes them try to do GPRS RA UPD without ever sending GPRS ATTACH first GPRS Session Management (04.08) implementation * Supports only one PDP Context per MS so far * Only IPv4, no IPv6 contexts yet * No persistent storage of active PDP contexts, i.e. all state is lost if process crashes or restarts * only supports one APN, routed to one single GGSN The TODO list in order of priority is: * Implement BSSGP flow control (per BSS and per MS) * Fix MM state machine issues, make sure inter-cell hand-over of GPRS works * Add access from SGSN to HLR (possibly using libosmo-{tcap,map}, maybe again a quick sqlite database hack meanwhile * Verify encrpytion support in LLC code * Work on header/payload compression I don't have a precise schedule yet for it. Anyone who wants to work on those issues is of course much welcome. Meanwhile, if you experience problems with GPRS, please take pcap files of the A-bis, Gb and GTP links and send them to me, with a detailed description of what goes wrong when. I'm happy to look at those issues and resolve them, if I get sufficiently verbose bug reports. Regards, Harald -- - Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)

1 0