[srslte-users] Srsue with USRP N310 for MIMO 2x2
Andre Puschmann
andre at srs.io
Mon Jun 21 07:54:31 UTC 2021
Hey,
On 18/6/21 9:54, François Légal wrote:
> By the way is there another way to avoid having to force the sampling
> rate like this ?
I wish there was. For many SDRs, including the Ettus B210 for example,
specifying the cell bandwidth isn't needed. It's read from the MIB that
is always sampled at 1.92Msps and then the SDR is retuned to match the
actual cell bandwidth. Unfortunately, this works better with some
devices than with others.
> e.g if we want the UE to attach to a 20 MHz cell we would have to change
> the sampling rate and restart srs.
You can always set it to 30.72 which is the maximum. srsUE will
down-sample on it's own.
Thanks
Andre
> Thanks a lot
>
> François
>
>
>> Le 16 juin 2021 à 22:41, Andre Puschmann <andre at srs.io
>> <mailto:andre at srs.io>> a écrit :
>>
>> Hey François,
>>
>>
>> [1]
>> https://docs.srsran.com/en/latest/app_notes/source/5g_nsa/source/index.html#srsue
>> <https://docs.srsran.com/en/latest/app_notes/source/5g_nsa/source/index.html#srsue>
>>
>> On 16/6/21 20:04, François Légal wrote:
>>> I’m trying to use srsue with a USRP N310 vs an LTE cell in conducted
>>> mode. Using RF port 0 and 1 on the same daughter board.
>>>
>>> -When the cell is configured in SISO, the srsue can attach without any
>>> issue.
>>> -When the cell is configured in 2x2 MIMO, srsue cannot attach at all.
>>> Srsue is configured with 2 antennas in ue.conf.
>>
>> that is likely an issue with UHD and how the streamers are configured.
>> You can circumvent that by configuring a fixed sample rate that is
>> configured at startup, see [1] for how to do that with the UE.
>>
>> For the N310, use 15.36e6 as sample rate for 10 MHz cells. And make sure
>> to set the LTE sample rate option.
>>
>>>
>>> Based on the logs, srsue can decode the MIB on the small band of 1920.
>>> Then there is a switch to a wider band of 15360 (for 10 MHz chBW cell),
>>> the MIB is not decoded anymore.
>>> It seems that there is no correlation peak found for PSS this time in
>>> this second stage for cell search / CAMPING
>>> Then we are stuck in this SYNC attempt loop.
>>>
>>> With a USRP B210, the attach works fine in MIMO 2x2
>>>
>>> The call to srsue is made without specific RF options in the command
>>> line. All other options are specified in the ue.conf file.
>>
>> try:
>>
>> [rf]
>> nof_antennas = 2
>> device_name = uhd
>> device_args = type=x300,clock=external,sampling_rate=15.36e6
>> srate = 15.36e6
>>
>> [expert]
>> lte_sample_rates = true
>>
>>>
>>> 1-Is it a known issue and it there an easy workaround such a using
>>> specific RF parameters for USRP ?
>>
>> See above.
>>
>>> 2-To which extend srsue (srsran) has been validated in front of a USRP
>>> N310 in MIMO 2x2, 4x4 etc.
>>
>> We don't use the N310 a lot and it also doesn't work well with UHD4. I
>> strongly recommend to use 3.15.
>>
>> In general, we don't support 4x4 MIMO, "just" 2x2.
>>
>>
>> Cheers
>> Andre
>
--
Andre Puschmann
Software Radio Systems (SRS)
https://www.srs.io
andre at srs.io
PGP/GnuPG key: 7F4858D408EA231B
fingerprint: 62C4 EE1C 69CE 9086 FB75 168A 7F48 58D4 08EA 231B
More information about the srsran-users
mailing list