[srslte-users] SIB1 decoding unreliable

[Balthasar Indermuehle]

Hi all,

I'm mapping my local 4G environment using srsLTE with an intel i7 quad core
laptop and a pair of Ettus B200 USB3 SDRs - one is cycling through the
available bands finding MIBs, the other is taking the frequencies it found
MIBs at and tries extracting a SIB1 to get the cell details. I'm
successfully decoding MIBs for all known operators in the area on all the
right frequencies. But the SIB1 messages on many frequencies take a very
long time until one arrives that is successfully decoded, and for some
frequencies, even with excellent signal strength reported (PSS power of >
-10 dB), srsue is simply never decoding a SIB1 (but decodes tons of MIB
frames while looking for the SIB1).

I was expecting the SIB1 to be transmitted on a similar schedule as the
MIB, so am a little baffled as to why there would be such a huge difference
in decoding success.

But I'm also seeing a lot of decoded MIBs on frequencies that are offset by
one channel, so evidently the MIBs are much easier to decode than a SIB and
don't even require a correct center frequency.

I'm running srsue with root privileges so it can run with realtime
priority. There seems to be no difference in decoding success (or lack
thereof) whether srsue is running concurrently with cell_search or not.

Is this a decoding performance issue? Or are SIBs only transmitted under
certain conditions and I just happen to miss those?

Any hints/ideas as to why SIB1 decoding performance is so poor would be
much appreciated.

Cheers

Balthasar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.srsran.com/pipermail/srsran-users/attachments/20201217/f5210195/attachment.htm>