[srslte-users] Bug with PUSCH (Collision) in srsENB

Ilshat Shaidullin ilshat.desu at gmail.com
Fri Nov 16 08:58:37 UTC 2018


Hello SRS Team,

I've noticed strange behavior in srsENB. It occurs when I connect 2 UEs.
I've attached log files (info level for phy, mac and rrc).

Let the 1st UE be A and the 2nd UE be B.
[FirstExample.log] I successfully connect A with rnti=0x46. Then I try to
connect B with rnti=0x47. But after eNB sends RRCConnectionSetup all UL
transmissions of B are always KO (though SNR is good). B can be
successfully connected only with next rnti (0x48).

[SecondExample.log] I successfully connect B with rnti=0x46. Then I
successfully connect A with rnti=0x47. But as soon as A is connected,  B's
UL transmissions are always KO (though SNR is good). So after some time B
will perform RACH and connect with new rnti. Only after that B's UL
transmissions will be OK.


I think that in the examples above when eNB sends DCI Format 0 for B, A
mistakenly uses it and performs UL transmission too. This may happen
because C-RNTI values are close to each other. If I change C-RNTI
allocation method ( replace "last_rnti++"  with "last_rnti += 10" in MAC )
I don't observe this issue.

What can cause such strange behavior?


Best regards,
Ilshat
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.srsran.com/pipermail/srsran-users/attachments/20181116/7422dae2/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Logs.zip
Type: application/zip
Size: 3479966 bytes
Desc: not available
URL: <https://lists.srsran.com/pipermail/srsran-users/attachments/20181116/7422dae2/attachment.zip>


More information about the srsran-users mailing list