- OpenBSC - lists.osmocom.org

WG: Why TDM time slot time duration is 577us ??
by Andreas.Eversberg 09 Jan '11

09 Jan '11

in addition: Traffic Multiframe Structures - The 26 traffic multiframe structure is used to send information on the traffic channel. The 26 traffic multiframe structure is used to combine user data (traffic), slow control signaling (SACCH), and idle time period. The idle time period allows a mobile device to perform other necessary operations such as monitoring the radio signal strength level of a beacon channel from other cells. The time interval of a 26 frame traffic multiframe is 6 blocks of speech coder data (120 msec). (http://www.althos.com/tutorial/GSM-tutorial-frame-structure.html) one encoded speech block lasts 20ms. 576,92307692307692307692307692308us per slot 4615,3846153846153846153846153846 per 8 slots (1 frame) 120ms per 26 frames (6 speech blocks)

4 4

Unknown issue with GGSN/SGSN
by Chris Rankine 08 Jan '11

08 Jan '11

I am trying to implement an experimental EDGE deployment using one of the DCS1800 models. (If needed for debugging purposes, I can request access to a GSM900 for GPRS only testing). I have everything compiled and running with the newest builds as of a few or two ago. EDGE works most of the time, however randomly it stops working. I have the SGSN and GGSN running, and typically during these down periods I will see messages appearing on the SGSN, but none of the typical traffic flowing messages on the GGSN (typical messages being "Received packet from tun!" and "encaps_tun. Packet received: forwarding to tun"). Typically if I wait a period of time, sometimes about 2 minutes, other times for a long time, then the connection returns. During this entire time the phone still displays the "E" meaning it believes it has an EDGE connection. If I reboot the phone I will get the PDP entry deleted and requested messages. As for the SGSN, I am new to GPRS and EDGE and I am not familiar with a lot of the terms that are appearing, so I can not comment on what is good or bad. I get the same results if I put three IPs on one VM and have it running OpenBSC, OpenGGSN, and the OpenBSC based SGSN, as compared to when I give one VM to OpenBSC and another dual homed VM hosts OpenGGSN and SGSN. I guess what I am asking is what do you suggest I should be looking into to try to figure out what is causing this problem? I have two old Android phones to play with.. So, at least for the Android phones I have better access at debugging issues on the phone. Attached below are the relevant parts of the GGSN conf, SGSN conf, and openBSC conf. Thanks for any suggestions. Chris Rankine ##### /etc/ggsn.conf fg debug listen 192.168.21.189 pcodns1 130.207.244.251 pcodns2 130.207.244.251 ###### osmo_sgsn.cfg sgsn gtp local-ip 192.168.21.188 ggsn 0 remote-ip 192.168.21.189 ggsn 0 gtp-version 1 ns timer tns-block 3 timer tns-block-retries 3 timer tns-reset 3 timer tns-reset-retries 3 timer tns-test 30 timer tns-alive 3 timer tns-alive-retries 10 encapsulation udp local-ip 192.168.21.188 encapsulation udp local-port 23000 encapsulation framerelay-gre enabled 0 bssgp ##### openbsc.cfg network network country code 1 mobile network code 1 short name OpenBSC long name OpenBSC auth policy closed location updating reject cause 13 encryption a5 0 neci 1 rrlp mode none mm info 1 handover 0 handover window rxlev averaging 10 handover window rxqual averaging 1 handover window rxlev neighbor averaging 10 handover power budget interval 6 handover power budget hysteresis 3 handover maximum distance 9999 timer t3101 10 timer t3103 0 timer t3105 0 timer t3107 0 timer t3109 0 timer t3111 0 timer t3113 60 timer t3115 0 timer t3117 0 timer t3119 0 timer t3141 0 bts 0 type nanobts band DCS1800 cell_identity 0 location_area_code 1 training_sequence_code 7 base_station_id_code 63 ms max power 40 cell reselection hysteresis 4 rxlev access min 0 channel allocator ascending rach tx integer 9 rach max transmission 7 ip.access unit_id 3 0 oml ip.access stream_id 255 gprs mode egprs gprs routing area 1 gprs cell bvci 2 gprs nsei 1 gprs nsvc 0 nsvci 1 gprs nsvc 0 local udp port 23000 gprs nsvc 0 remote udp port 23000 gprs nsvc 0 remote ip 192.168.21.188 trx 0 rf_locked 0 arfcn 514 nominal power 70 max_power_red 0 rsl e1 tei 0 timeslot 0 phys_chan_config CCCH+SDCCH4 timeslot 1 phys_chan_config SDCCH8 timeslot 2 phys_chan_config TCH/F timeslot 3 phys_chan_config PDCH timeslot 4 phys_chan_config TCH/F timeslot 5 phys_chan_config PDCH timeslot 6 phys_chan_config PDCH timeslot 7 phys_chan_config PDCH

2 1

Re: OpenBSC on Solaris (WAS: Re: Notes from the 27C3 network)
by omar.atia＠its.ws 07 Jan '11

07 Jan '11

Dears, Sorry for hijacking the thread , next time I will create new thread ... I will try to bind using IP to avoid using SO_BINDTODEVICE ,and I will google IP_BOUND_IF as well. Another issue is the linux command for vty which is the c function under linux called cfmakeraw which I removed from vty.c cause there is no library under SUN for that command or function to link with. Can you please ensure that this commented line will not affect vty functionality ? Thanks, Omar atia ------Original Message------ From: Holger Hans Peter Freyther Sender: openbsc-bounces(a)lists.gnumonks.org To: openbsc(a)lists.gnumonks.org Subject: OpenBSC on Solaris (WAS: Re: Notes from the 27C3 network) Sent: Dec 31, 2010 11:28 AM On 12/31/2010 12:30 AM, Omar Atia wrote: > Dears ,, please help :: Dear Omar, please do not hijack threads, start a new one instead. > > in ipaccess-find.c , can we use bind instead of setsockopt cause solaries > doesn't have SO_BINDTODEVICE defined in socket.h. For ipaccess-find is sending a broadcast packet, so it is important to select the physical interface it should be send on. For the user it is more convenient to write the interface name then to remember the ip address of this interface. Does Solaris provide another way to bind to a specific interface? What about IP_BOUND_IF? If that is not the case we could consider doing a normal bind and bind to the interface via an ip... z. Sent via BlackBerry® from mtc touch

4 12

AW: Why TDM time slot time duration is 577us ??
by Andreas.Eversberg 07 Jan '11

07 Jan '11

576,92307692307692307692307692308us per slot 4615,3846153846153846153846153846 per 8 slots (1 frame) 120ms per 26 frames ________________________________ Von: openbsc-bounces(a)lists.gnumonks.org [mailto:openbsc-bounces@lists.gnumonks.org] Im Auftrag von Nataraju A B Gesendet: Freitag, 7. Januar 2011 13:46 An: openbsc(a)lists.gnumonks.org Betreff: Why TDM time slot time duration is 577us ?? Hi All, I wish I am posting this query to the right mailing list. If not please let me know the correct list for this query. I understand the basics of GSM and wireless in general. I have a very basic question. Why did the time slot value considered to be 577us ? I searches Google for some useful doc, which clarifies the reason behind selection of time slot ? Thanks, Nataraju A B

1 0

Why TDM time slot time duration is 577us ??
by Nataraju A B 07 Jan '11

07 Jan '11

Hi All, I wish I am posting this query to the right mailing list. If not please let me know the correct list for this query. I understand the basics of GSM and wireless in general. I have a very basic question. Why did the time slot value considered to be 577us ? I searches Google for some useful doc, which clarifies the reason behind selection of time slot ? Thanks, Nataraju A B

1 0

AW: applying multibts-patch
by Andreas.Eversberg 07 Jan '11

07 Jan '11

you don't need that patch. you can configure the hfcmulti.ko using dmask and bmask: * dmask and bmask: * NOTE: One dmask value must be given for every HFC-E1 card. * If omitted, the E1 card has D-channel on time slot 16, which is default. * dmask is a 32 bit mask. The bit must be set for an alternate time slot. * If multiple bits are set, multiple virtual card fragments are created. * For each bit set, a bmask value must be given. Each bit on the bmask * value stands for a B-channel. The bmask may not overlap with dmask or * with other bmask values for that card. * Example: dmask=0x00020002 bmask=0x0000fffc,0xfffc0000 * This will create one fragment with D-channel on slot 1 with * B-channels on slots 2..15, and a second fragment with D-channel * on slot 17 with B-channels on slot 18..31. Slot 16 is unused. * If bit 0 is set (dmask=0x00000001) the D-channel is on slot 0 and will * not function. * Example: dmask=0x00000001 bmask=0xfffffffe * This will create a port with all 31 usable timeslots as * B-channels. * If no bits are set on bmask, no B-channel is created for that fragment. * Example: dmask=0xfffffffe bmask=0,0,0,0.... (31 0-values for bmask) * This will create 31 ports with one D-channel only. * If you don't know how to use it, you don't need it! hope this helps -----Ursprüngliche Nachricht----- Von: openbsc-bounces(a)lists.gnumonks.org [mailto:openbsc-bounces@lists.gnumonks.org] Im Auftrag von Thomas Ansorg Gesendet: Freitag, 7. Januar 2011 12:39 An: openBSC-Mailingliste Betreff: applying multibts-patch hello i downloaded kernelversion 2.6.30 from www.kernel.org for using the linux-2.6.30-hfcmulti-multibts.patch on it. unfortunately, some hunks in the c-file failed. where can i download a kernel that fits to that patch? T. -- Wer Rechtschreibfehler findet, darf sie behalten!

1 0

applying multibts-patch
by Thomas Ansorg 07 Jan '11

07 Jan '11

hello i downloaded kernelversion 2.6.30 from www.kernel.org for using the linux-2.6.30-hfcmulti-multibts.patch on it. unfortunately, some hunks in the c-file failed. where can i download a kernel that fits to that patch? T. -- Wer Rechtschreibfehler findet, darf sie behalten!

1 0

[ADM] OpenBSC git repository address change
by Harald Welte 07 Jan '11

07 Jan '11

Hi all! Today, I have moved the remaining projects on git://openbsc.osmocom.org/ (aka git://bs11-abis.gnumonks.org/) to the central git://git.osmocom.org/ location. This helps us to consolidate all our git repositories on one server. The git URL of the main OpenBSC repository has changed from git://bs11-abis.gnumonks.org/openbsc.git (read-only) gitosis@bs11-abis.gnumonks.org:openbsc.git (write) to git://git.osmocom.org/openbsc.git (read-only) gitosis@git.osmocom.org:openbsc.git (write) In order to update your local repository with the new URL, simply use the "git remote set-url" command, like in the following examples. for those of you with read-only access: git remote set-url origin git://git.osmocom.org/openbsc.git for people with write/commit access: git remote set-url origin gitosis@git.osmocom.org:openbsc.git The other two repositories that have changed in similar way are openggsn.git and chan_openbsc.git. Thanks for your attention, and sorry for any inconvenience, 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

Wireshark Patching...
by Omar Atia 06 Jan '11

06 Jan '11

Dear All, Please note that after getting the sources using git , I found wireshark directory contains patches ? I did some googling I found that if I need to patch wireshark I should run like the below command : 1. patch -p1 < patch path/name the configure and compile , but while checking wireshark sources I didn't find for packet-gsm_abis_oml.c in the path of wireshark /epan/dissectors? does the patch abis_oml.patch(available under wireshark directory with the delivery downloaded with git) create it ? Thanks, Omar Atia

1 0

IPA Stream IDs
by Holger Hans Peter Freyther 06 Jan '11

06 Jan '11

Hi all, I was very liberal in just using numbers from the 'open' number range. We should move to a different scheme and just take the liberty to take one ID and then have our own header inside there. I would propose to do something like this struct ipa_proto_hdr { uint16_t len; uint8_t proto; data[0]; } packed; struct osmo_ipa_proto_hdr { uint8_t osmo_proto; data[0]; } packed;. Any comments?

2 2