[srsran-users] srsUE not attaching to real eNB - bladeRF 2.0 micro xA5

[Antonio Albanese]

Thank you for your suggestion! It doesn't seem to solve it for me.

I used EARFCNs 100 and 1300. I tried wiggling around ±2 but with no luck
whatsoever. Do you remember what was the order of magnitude of this delta?

I tried the cell search. It turns out srsRAN finds the cells under test at
the correct EARFCNs even with the bladeRF. I don't know if that is good
information or not.

I also validated I only get RF overruns both with soapy and bladeRF
drivers. For instance, soapy reported around 3500 overruns just for the
cell search.

Thank you for your help!

On Sat 5. Nov 2022 at 17:16, Cagalli Athha <calgalli at gmail.com> wrote:

> Can you try to set different channels? Try to set Earfcn in enb.conf to
> be a channel frequency that is in round number. I have to same problem too.
> It seems that in some Earfcns frequency is off a bit and UE cannot
> connect. I use those  Earfcns in B210 and it seems to be ok. The problem
> is only for Bladerf. I guess it is a driver problem.
>
> On Sat, Nov 5, 2022 at 11:06 PM Antonio Albanese <
> antonioalbanese15 at gmail.com> wrote:
>
>> Hey fellas,
>>
>> I am having quite some trouble connecting to a real eNB using a bladeRF
>> 2.0 micro xA5.
>>
>> *Disclaimer*: when I use a USRP B210 with exactly the same ue.conf file,
>> srsue decodes MIB and SIBs, and attempts to attach. At the moment the
>> authentication fails but that is intended as I am using a random IMSI
>> (including the correct PLMN digits for my *desired* operator).
>>
>> On the other hand, whenever I use a bladeRF I cannot get any attach
>> attempt.
>>
>> I thought about clock instability issues, so I am using an external GPSDO
>> applied to the REFIN U.FL connector of the bladeRF. I checked and I can
>> successfully get the PLL to lock, so the VCTCXO should be super stable at
>> this point.
>>
>> I tried using both bladerf and soapy drivers. WIth the bladerf driver,
>> srsue reaches SYNC state CAMPING but then it detects a radio overflow and
>> tries to resync SFN in an infinite loop.
>> With soapy drivers, I have even worse luck since srsue never synchronizes
>> (this might be due to a missing setting for soapy indicating that the
>> bladeRF should use the REFIN clock signal, I have not looked into that yet).
>>
>> If the USRP did not work, I would think of some CPU bottleneck, which
>> would generate RF overflows. However, the USRP works perfectly and all my
>> problems appear only with the bladeRF.
>>
>> I am attaching here both debug logs, with soapy and bladerf drivers.
>>
>> Any suggestion is highly appreciated and would save a fellow srsRAN user
>> from desperation.
>>
>> Thank you so much!
>> Bye!
>> ᐧ
>>
> _______________________________________________
>> srsran-users mailing list
>> srsran-users at lists.srsran.com
>> https://lists.srsran.com/mailman/listinfo/srsran-users
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.srsran.com/pipermail/srsran-users/attachments/20221107/971068a5/attachment-0001.htm>