[srslte-users] Recieving further SIB messages

[Ismael Gomez]

Hi Enes,

For the CPU problems, make sure your CPU clock is fixed to the maximum and
the energy saving governor is disabled. You should be able to decode SIBs
with a standard dual-core 2GHz intel processor.

What I see from your logs is that the SNR is quite low ~0 dB, but the SIBs
are decoded correctly.

We have never decoded any SIBs other than SIB1 and SIB2. You are actually
the first one :).

I think that your problem has an easy fix. The current RRC implementation
stops the PDCCH search whenever a SIB2 is received. If the search window
for the SIB you are looking coincides with the SIB2 position then the PHY
will stop searching. So what you could try is to comment line 269 in rrc.c
so that the PHY keeps looking for the entire duration of the window.




On Thu, 26 May 2016 at 13:30 Enes ÜNAL <aenesunal at gmail.com> wrote:

> Hello Ismail,
>
> I just tried to get some logs, but I think it's almost impossible since my
> CPU is not a good one and intensive logging lead the UE to not catching any
> SIB1 message.
>
> Let me help you to understand the issue with something else,
>
> In SIB1, I see SIB2, SIB3, SIB5 messages are broadcasting with
> periodicities rf16, rf32, rf32 and the SI window size is read as 40.
>
> By that, I added some more `case`s to the switch in rrc::sib_search, to
> get SIB3, SIB4 and SIB5.
>
> There is no problem with SIB3, and since there is no SIB4 message
> currently broadcasting from E-UTRAN; I'm switching SIB5 message retrieval
> case.  For that, I'm calling mac_interface_rrc::bcch_start_rx with
> si_win_start [=sib_start_tti(tti, 32, 0)], and si_win_len
> [=si_window_length in SIB1].
>
> Examples
>
> ********************************************************
> tti: 6490, period: 32, siwinst: 6720, siwinlen: 40
> ********************************************************
> tti: 7055, period: 32, siwinst: 7360, siwinlen: 40
> ********************************************************
> tti: 7142, period: 32, siwinst: 7360, siwinlen: 40
> ********************************************************
> tti: 7166, period: 32, siwinst: 7360, siwinlen: 40
> ********************************************************
>
> Also I tried to decode the messages retrieved by rrc::write_pdu_bcch_dlsch
> with compiled asn's. For that at the start of this method I added just a
> couple of lines to get the data.
>
> FILE *f = fopen(fname, "w");
> fwrite(pdu->buffer, 1, pdu->N_bytes, f);
>
> I'm attaching some of these files to this mail. As I decoded, I cannot see
> any SIB5 messages in these recorded raw data. I've used the procedure
> explained in the following link:
> http://sdr-x.github.io/LTE-SIB-decoding-by-asn1c/
>
> I tried this with 8 different frequencies and ~30 different physical cell
> ids. Could not get just one SIB5 message.
>
> The first thing I want to know if you can get any SIB4, SIB5 etc. messages
> before?
>
> The second one I want to ask about is what is the system requirements to
> run the UE flawlessly?
>
> Thank you for your interest and support,
> Best,
> Enes.
>
> 2016-05-25 19:59 GMT+03:00 Ismael Gomez <
> ismael.gomez at softwareradiosystems.com>:
>
>> Hi Enes,
>>
>> your approach is correct. Can you provide the logs of your execution? Set
>> the log level to info for all layers in the ue.conf file.
>>
>> thanks
>>
>> On Wed, 25 May 2016 at 08:49 Enes ÜNAL <aenesunal at gmail.com> wrote:
>>
>>> Hello,
>>>
>>> First I should thank for all of the great work done by SRS.
>>>
>>> I'm SDR hobbyist student and currently experimenting srsUE. When I went
>>> into deep, and read about the LTE standard specs, learnt that other System
>>> Information Broadcast Messages is also broadcasting from the real LTE
>>> networks. Then I decided to get some of it. As I see, SIB1 and SIB2 are
>>> currently being read by srsUE RRC.
>>>
>>> From there, I could get the SchedulingInfoList from SIB1 message. And
>>> saw that eNB is set to broadcast SIB2, SIB3, SIB5, etc. Then with some
>>> additions, I tried to get other SIB messages. This work lead me to
>>> successfully get SIB3 messages with the same window with SIB2. However when
>>> I try to get SIB5 message, even if I
>>> called mac_interface_rrc::bcch_start_rx with correct windows start and
>>> length; there is no SIB5 message type is retrieved.
>>>
>>> Did anyone try it before, or it's not tested? Should I go deep into
>>> liblte_rrc?
>>>
>>> Thank you,
>>> Best,
>>> Enes.
>>> _______________________________________________
>>> srslte-users mailing list
>>> srslte-users at lists.softwareradiosystems.com
>>> http://www.softwareradiosystems.com/mailman/listinfo/srslte-users
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.srsran.com/pipermail/srsran-users/attachments/20160526/de485eab/attachment.htm>