[srslte-users] Commercial UE Attach problems

Fri Nov 17 08:47:19 UTC 2017

Just a quick update on the matter: The problem was the TAC configuration in srsENB not matching with EPC.
After that, attachment was possible and sucessful. Now I am working on stabilizing connection. It seems that data plane is not working and after some time UE gets released again.
But thank you for the help so far.

Regards,
Jens

Von: Ismael Gomez [mailto:ismael.gomez at softwareradiosystems.com]
Gesendet: Mittwoch, 15. November 2017 10:47
An: Pilz, Jens <jens.pilz at hhi.fraunhofer.de>
Cc: srslte-users at lists.softwareradiosystems.com
Betreff: Re: [srslte-users] Commercial UE Attach problems

Looks like the EPC rejected the UE. For the pcap you'll need Wireshark MAC-LTE dissector

On Wed, 15 Nov 2017 at 10:35 Pilz, Jens <jens.pilz at hhi.fraunhofer.de<mailto:jens.pilz at hhi.fraunhofer.de>> wrote:
Here you go, but the pcap looks funny to me :-D

--
Dipl.-Phys. Jens Pilz
Project Manager

Fraunhofer Heinrich Hertz Institute HHI
Wireless Communications and Networks Department
Einsteinufer 37, 10587 Berlin, Germany<https://maps.google.com/?q=Einsteinufer+37,+10587+Berlin,+Germany&entry=gmail&source=g>
phone + 49 30 31002 390<tel:+49%2030%2031002390> | mobile + 49 179 1497494<tel:+49%20179%201497494>
email jens.pilz at hhi.fraunhofer.de<mailto:jens.pilz at hhi.fraunhofer.de>
www.hhi.fraunhofer.de<http://www.hhi.fraunhofer.de> | Facebook<https://www.facebook.com/FraunhoferHHI1928> | Twitter<https://twitter.com/FraunhoferHHI>

Von: Ismael Gomez [mailto:ismael.gomez at softwareradiosystems.com<mailto:ismael.gomez at softwareradiosystems.com>]
Gesendet: Mittwoch, 15. November 2017 10:19

An: Pilz, Jens <jens.pilz at hhi.fraunhofer.de<mailto:jens.pilz at hhi.fraunhofer.de>>
Cc: srslte-users at lists.softwareradiosystems.com<mailto:srslte-users at lists.softwareradiosystems.com>
Betreff: Re: [srslte-users] Commercial UE Attach problems

yeah sorry, I meant /tmp/enb.log or /tmp/enb.pcap.

On Wed, 15 Nov 2017 at 10:18 Pilz, Jens <jens.pilz at hhi.fraunhofer.de<mailto:jens.pilz at hhi.fraunhofer.de>> wrote:
I assume these files are generated by srsUE?

I did not yet test with srsUE but with commercial USB LTE dongle.

Give me some time and I will test with srsUE. (need to figure out how to set IMSI, etc)

Regards,
Jens

Von: Ismael Gomez [mailto:ismael.gomez at softwareradiosystems.com<mailto:ismael.gomez at softwareradiosystems.com>]
Gesendet: Mittwoch, 15. November 2017 10:04

An: Pilz, Jens <jens.pilz at hhi.fraunhofer.de<mailto:jens.pilz at hhi.fraunhofer.de>>
Cc: srslte-users at lists.softwareradiosystems.com<mailto:srslte-users at lists.softwareradiosystems.com>
Betreff: Re: [srslte-users] Commercial UE Attach problems

Hi Jens,

Can you attach the /tmp/ue.log and /tmp/ue.pcap (set log level to info and enable pcap in ue.conf).

Thanks

On Wed, 15 Nov 2017 at 08:55 Pilz, Jens <jens.pilz at hhi.fraunhofer.de<mailto:jens.pilz at hhi.fraunhofer.de>> wrote:
Hi Ismael,

comments inline.

Von: Ismael Gomez [mailto:ismael.gomez at softwareradiosystems.com<mailto:ismael.gomez at softwareradiosystems.com>]
Gesendet: Dienstag, 14. November 2017 14:35
An: Pilz, Jens <jens.pilz at hhi.fraunhofer.de<mailto:jens.pilz at hhi.fraunhofer.de>>
Cc: srslte-users at lists.softwareradiosystems.com<mailto:srslte-users at lists.softwareradiosystems.com>
Betreff: Re: [srslte-users] Commercial UE Attach problems

1)      You can not set nof_phy_threads bigger than 3. Only 1, 2 o 3 are valid.
[Pilz, Jens] Okay changed back to 3.

2)      I think the TX/RX offset is not properly calibrated for the X300. Can you try setting rf.time_adv_nsamples=18

[Pilz, Jens] Thank you fort he tip. We did additional testing with different values and also used the PRACH calibration function from srsLTE. (which gave 65)…but it seems with 18 the rnti spamming stopped.

3)      Your signal does not look very good. Make sure TX and RX antennas are orthogonal and correct for your band

[Pilz, Jens] The signal by eNodeB looks good with monitor antenna and spectrum analyser with LTE software.
I don’t trust the X310 with UBX doughter to generate a good signal (at least on TX with high gain the signal gets distorted quite fast) I am playing with various gain settings now, but so far the attach wasn’t successful
                        Anyhow, the messages in EPC look suspicious to me and something might be wrong there.

4)      Make sure you adjust the PUSCH/PUCCH power control settings in sib.conf according to your TX power

[Pilz, Jens] Okay will adapt that.

On Tue, 14 Nov 2017 at 14:30 Pilz, Jens <jens.pilz at hhi.fraunhofer.de<mailto:jens.pilz at hhi.fraunhofer.de>> wrote:
Hi all,

i got everything working on a basic level: OpenEPC 6 -> srsLTE -> USRP x310 -> Antenna -> Rohde & Schwarz FSW13
When I try to attach commercial UE with working sim card (setup is verified with different eNB on same core network),
I see a lot of rnti created in srsENB:
RACH:  tti=7321, preamble=17, offset=0, temp_crnti=0x46
RACH:  tti=7411, preamble=13, offset=0, temp_crnti=0x47
SCHED: Could not transmit RAR within the window (RA TTI=7411, Window=10, Now=7428)
RACH:  tti=7431, preamble=28, offset=0, temp_crnti=0x48
Disconnecting rnti=0x47.
RACH:  tti=7521, preamble=2, offset=0, temp_crnti=0x49
SCHED: Could not transmit RAR within the window (RA TTI=7521, Window=10, Now=7541)
RACH:  tti=7541, preamble=14, offset=0, temp_crnti=0x4a
Disconnecting rnti=0x49.
RACH:  tti=7631, preamble=38, offset=0, temp_crnti=0x4b
Disconnecting rnti=0x46.
Disconnecting rnti=0x48.
Disconnecting rnti=0x4b.
Disconnecting rnti=0x4a.
RACH:  tti=8451, preamble=21, offset=0, temp_crnti=0x4c
Disconnecting rnti=0x4c.
RACH:  tti=9931, preamble=38, offset=0, temp_crnti=0x4d
Disconnecting rnti=0x4d.
RACH:  tti=10021, preamble=38, offset=0, temp_crnti=0x4e
Disconnecting rnti=0x4e.
RACH:  tti=10001, preamble=26, offset=0, temp_crnti=0x4f
Disconnecting rnti=0x4f.
RACH:  tti=1251, preamble=26, offset=0, temp_crnti=0x50
Disconnecting rnti=0x50.

This can be easily reproduced and linked to commercial LTE USB Dongle (Huawei E392).
I also attached eNB log.
In my opinion, one UE should only create one RACH?!?

Also there is some misbehavior in MME, but I thing this is correlated to so many “RNTI” created.

Regards,
Jens

--
Dipl.-Phys. Jens Pilz
Project Manager

Fraunhofer Heinrich Hertz Institute HHI
Wireless Communications and Networks Department
Einsteinufer 37, 10587 Berlin, Germany<https://maps.google.com/?q=Einsteinufer+37,+10587+Berlin,+Germany&entry=gmail&source=g>
phone + 49 30 31002 390<tel:+49%2030%2031002390> | mobile + 49 179 1497494<tel:+49%20179%201497494>
email jens.pilz at hhi.fraunhofer.de<mailto:jens.pilz at hhi.fraunhofer.de>
www.hhi.fraunhofer.de<http://www.hhi.fraunhofer.de> | Facebook<https://www.facebook.com/FraunhoferHHI1928> | Twitter<https://twitter.com/FraunhoferHHI>

_______________________________________________
srslte-users mailing list
srslte-users at lists.softwareradiosystems.com<mailto: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/20171117/c5a587b0/attachment.htm>