Hi Alexander,

Thats what I was afraid of :-)

Anyway, even having EDGE with GMSK MCSs is a huge step forward, so I can't say anything but thank you, and I hope I can contribute with some testing and protocol analysis. :-)

Now we are using OpenBSC in two lab exercises and I hope to extend it with SDR and EDGE. I can't wait to try it :-)

Regards, Csaba

----- Eredeti üzenet ----- Feladó: "Alexander Chemeris" alexander.chemeris@gmail.com Címzett: "Sipos Csaba" sipos.csaba@kvk.uni-obuda.hu Másolatot kap: "OpenBSC Mailing List" openbsc@lists.osmocom.org, osmocom-net-gprs@lists.osmocom.org, "Jacob Erlbeck" jerlbeck@sysmocom.de Elküldött üzenetek: Csütörtök, 2016. Február 4. 23:25:30 Tárgy: Re: EDGE over osmo-pcu

Hi Sipos,

Osmo-trx does not support 8-PSK at this moment, so you should implement it if you want to test it. ;)

Please excuse typos. Written with a touchscreen keyboard.

-- Regards, Alexander Chemeris CEO Fairwaves, Inc. https://fairwaves.co On Feb 4, 2016 1:43 PM, "Sipos Csaba" sipos.csaba@kvk.uni-obuda.hu wrote:

Hi Jacob,

Congratulations and thank you for all your hard work.

I hope I can do some testing on the weekend with a B200. I will also update the wiki article if needed.

Just one question: for MCS5-9 in the downlink there is no need to update the Osmo-TRX code? Does it support 8PSK as it is now?

Regards, Csaba

----- Eredeti üzenet ----- Feladó: "Jacob Erlbeck" jerlbeck@sysmocom.de Címzett: osmocom-net-gprs@lists.osmocom.org Másolatot kap: openbsc@lists.osmocom.org Elküldött üzenetek: Csütörtök, 2016. Február 4. 18:59:05 Tárgy: EDGE over osmo-pcu

Hi everybody,

we have just done the first download of an Internet page using the current bleeding EDGE development version of the OSMO-PCU. Supported are MCS-1 to MCS-9 in downlink and MCS-1 to MCS-4 in uplink direction.

The download of 'www.osmocom.org' has been documented in a PCAP file and can be downloaded from

https://openbsc.osmocom.org/trac/wiki/osmo-pcu

To decode the data blocks, an experimental patch for the gsmtab dissector is also available from there.

Cheers

Jacob

--

• Jacob Erlbeck jerlbeck@sysmocom.de http://www.sysmocom.de/

=======================================================================

• sysmocom - systems for mobile communications GmbH
• Alt-Moabit 93
• 10559 Berlin, Germany
• Sitz / Registered office: Berlin, HRB 134158 B
• Geschaeftsfuehrer / Managing Directors: Holger Freyther, Harald Welte

Hi Sipos

On 04.02.2016 22:40, Sipos Csaba wrote:

Congratulations and thank you for all your hard work.

Thank you. And many thanks to Sysmocom and On-Waves to make this possible.

I hope I can do some testing on the weekend with a B200. I will also update the wiki article if needed.

It is not in master yet and my development (wip) branch is still a mess (GPRS is b0rken in the latter), but this should improve soon. I'll put the revised commits to jerlbeck/master first.

Just one question: for MCS5-9 in the downlink there is no need to update the Osmo-TRX code? Does it support 8PSK as it is now?

I've only tested it with the sysmobts with direct DSP access so far and haven't looked at the Osmo-TRX or the pcu-if socket based connection to the BTS. But currently the CS/MCS choice is only communicated indirectly via the RLC block size which is unique for each coding scheme. The CPS (coding and puncturing scheme) is also not passed explicitly, only within the header which is different for each MCS header type. So the API for downlink blocks hasn't change, and it should only depend of the MCS encoding/decoding capabilities of the RF DSP/software.

Regards

Jacob

----- Eredeti üzenet ----- Feladó: "Jacob Erlbeck" jerlbeck@sysmocom.de Címzett: osmocom-net-gprs@lists.osmocom.org Másolatot kap: openbsc@lists.osmocom.org Elküldött üzenetek: Csütörtök, 2016. Február 4. 18:59:05 Tárgy: EDGE over osmo-pcu

Hi everybody,

we have just done the first download of an Internet page using the current bleeding EDGE development version of the OSMO-PCU. Supported are MCS-1 to MCS-9 in downlink and MCS-1 to MCS-4 in uplink direction.

The download of 'www.osmocom.org' has been documented in a PCAP file and can be downloaded from

https://openbsc.osmocom.org/trac/wiki/osmo-pcu

To decode the data blocks, an experimental patch for the gsmtab dissector is also available from there.

Cheers

Jacob

Hi Jacob,

Thanks for the information.

Just wanted to test the new code, and I hit this problem:

make[1]: Entering directory `/root/new_osmocom/osmo-pcu/src' CXX decoding.lo decoding.cpp:26:34: fatal error: osmocom/core/bitcomp.h: No such file or directory #include <osmocom/core/bitcomp.h> ^ compilation terminated. make[1]: *** [decoding.lo] Error 1 make[1]: Leaving directory `/root/new_osmocom/osmo-pcu/src' make: *** [all-recursive] Error 1

The libosmocore (master) is updated today, and the missing file is actually in /usr/local/include/osmocom/core/bitvec.h

I tried both "jerlbeck/testing/pcu" and "jerlbeck/master" branches, the error is the same.

Maybe I am missing something?

Regards, Csaba

----- Eredeti üzenet ----- Feladó: "Jacob Erlbeck" jerlbeck@sysmocom.de Címzett: "Sipos Csaba" sipos.csaba@kvk.uni-obuda.hu Másolatot kap: osmocom-net-gprs@lists.osmocom.org, openbsc@lists.osmocom.org Elküldött üzenetek: Kedd, 2016. Február 9. 23:28:24 Tárgy: Re: EDGE over osmo-pcu

Hi Sipos,

On 05.02.2016 11:22, Jacob Erlbeck wrote:

On 04.02.2016 22:40, Sipos Csaba wrote:

...

I hope I can do some testing on the weekend with a B200. I will also update the wiki article if needed.

It is not in master yet and my development (wip) branch is still a mess (GPRS is b0rken in the latter), but this should improve soon. I'll put the revised commits to jerlbeck/master first.

GPRS is fixed and jerlbeck/master is updated. Make sure to limit the MCS according to the HW/SW, so use the 'mcs max 4' command if the HW does not support 8PSK. Note that 8PSK is only supported for downlink yet, but the 'mcs max' command does not check this. So you should use 'mcs max 9 4' explicitely (this should be fixed by the additional commits (currently) in jerlbeck/testing/pcu, which adds that check and proper default values).

Jacob