NB-IOT JIO SIM Not connecting asset tracker Work Up to RRC Connected and stuck there.

[Hardik.Harpal]

Using NB- IOT JIO SIM in india, not connecting asset tracker Work Up to RRC Connected and stuck there. This is custom code very similar to asset tracker example. We are using custom board, same setup works fine on Ibasis SIM in USA.

[00:00:00.215,576] [0m<inf> asset_tracker: Connecting to LTE network.[0m
[00:00:00.215,942] [0m<inf> asset_tracker: This may take several minutes.[0m
[00:00:00.223,205] [0m<dbg> lte_lc.lte_lc_system_mode_set: Sending AT command to set system mode: AT%XSYSTEMMODE=0,1,1,0[0m
[00:00:00.259,643] [0m<inf> lte_lc: PDP Context: AT+CGDCONT=0,"IPV6","JioIPIoT"[0m
[00:00:00.267,608] [0m<dbg> lte_lc.lte_lc_system_mode_set: Sending AT command to set system mode: AT%XSYSTEMMODE=0,1,1,0[0m
[00:00:27.635,711] [0m<dbg> lte_lc.at_handler: +CEREG notification: +CEREG: 2,"C061","00282218",9
[0m
[00:00:27.636,444] [0m<inf> asset_tracker: LTE cell changed: Cell ID: 2630168, Tracking area: 49249[0m

trace-2021-07-13T00-06-58.563Z.pcapng

trace-2021-07-13T00-06-58.563Z.bin

[Heidi]

Hi! Håkon has gone on vacation so I have taken over this case.

The conclusion from the modem trace you provided is that the Attach procedure is prevented due to the modem reset loop restriction feature which was introduced in MFW 1.3.0. We have just released a white paper on the feature which should explain the functionality in detail: nWP042 - nRF9160 Modem Reset Loop Restriction

In short, the feature restricts Attach attempts to the network for 30 minutes in the case where the modem detects that the UE is in a reset loop. This can also happen if the modem is manually reset too many times.

The AT+CFUN = 0 command should be used for shutting down the modem gracefully in order to avoid this restriction. 

Please let me know if this solves the issue you are seeing. 

Best regards,

Heidi

[Hardik.Harpal]

Hi,

But I am seeing similar results in Modem firmware version 1.2.3 and earlier version on Dev kit I have 1.2.3 let me send you modem log from it. I will use MQTT simple example and configure it for nbiot.

Can AT_Client example running with link monitor can provide modem trace by default or do I have to change prj config changes to enable it?

Can you analyze this  new log withing next 22 hours? I have meeting with  JIO support team. If anything from their side I can provide our observation from Log they're asking for more detailed modem log as I explained in previous replies, I shared same modem trace (wireshark playable) with them. There are higher chances it is from network, but we need to figure it out as it is a month long now.

Regards,

Hardik Harpal

[Heidi]

Hardik.Harpal said:

Can AT_Client example running with link monitor can provide modem trace by default or do I have to change prj config changes to enable it?

 The AT Client sample does not have the correct config enabled by default, you have to add CONFIG_NRF_MODEM_LIB_TRACE_ENABLED=y to the prj.conf file.

Hardik.Harpal said:

Can you analyze this  new log withing next 22 hours?

 I am dependent on the schedule of the modem team, but if you provide me the trace within the next hour they should have enough time to analyze it by tomorrow morning. 

Unfortunately, the converted modem logs are only for use internally in Nordic so I can't provide the converted file, but I will try to provide the relevant information (LTE Physical, MAC, RRC, NAS signaling).

Best regards,

Heidi

[Hardik.Harpal]

Application: mqtt_simple, modem FW:1.2.3, board: nrf91DK, NCS1.6.0

5808.mqtt_simple.rar

Application Log:

16:01:22.338, *** Booting Zephyr OS build v2.6.0-rc1-ncs1 ***
16:01:23.337, [00:00:00.212,249] [0m<inf> mqtt_simple: The MQTT simple sample started[0m
16:01:23.448, [00:00:00.218,902] [0m<inf> mqtt_simple: Disabling PSM and eDRX[0m
16:01:23.448, [00:00:00.219,696] [0m<inf> mqtt_simple: LTE Link Connecting...[0m
16:02:17.323, +CEREG: 2,"C057","00094218",9,0,0,"11100000","11100000"
16:02:17.445, +CSCON: 1
16:02:17.666, +CEREG: 0,"C057","00094218",9,0,8,"11100000","11100000"
16:02:17.666, +CSCON: 0

trace-2021-08-03T10-44-01.054Z.pcapng

trace-2021-08-03T10-44-01.054Z.bin

[Hardik.Harpal]

Heidi Any update?

[Heidi]

There are a lot of traces lost during decoding, which isn't something we've seen before. Unfortunately, our tracing tool experts aren't back to work until next week.

The last NAS level trace shows that it is requesting RRC connection establishment for the Attach Request which was not the case in MFW 1.3.0 in which Attach procedure was blocked due to reset loop restriction.

[Hardik.Harpal]

Okay. So let's wait for next week, If anything required again, just let me know.

If any progress made from my side I will update you.

[Hardik.Harpal]

Heidi said:

There are a lot of traces lost during decoding, which isn't something we've seen before. Unfortunately, our tracing tool experts aren't back to work until next week.

If I collect log from old trace collector, I am using trace collector v2 preview v0.0.10, is it possible to analyze log without tracing tool expert.  I am thinking may be issue with this preview version and old modem firmware version 1.2.3

[Heidi]

Hi again,

So we had the tracing tool experts take a look at the traces you provided.

The MFW 1.3.0 log seems to be ok from a tracing perspective, i.e. no errors. So if you could get a new log with 1.3.0 starting from startup and about 1 minute long. Please confirm that the reset loop detection is not active before taking the log. Modem domain event notifications, including notifications of the modem reset loop restriction's activation, can be subscribed with AT%MDMEVAT=1.

With regard to the 1.2.3 log there are a lot of errors, and we suspect the problem lies in NCS. 

Which version of NCS are you using with MFW 1.2.3, and which version are you using with MFW 1.3.0?

Best regards,

Heidi

[Hardik.Harpal]

NCS v1.5.1 in both cases. 
Just for information same setup works in the USA(around Silicon Valley) with iBasis SIMs on both NBIoT and CAT-M networks.

I will provide log, but previous log was similar from starting and 1 minute long.

I will do AT+CFUN=0 before running this firmware. 

[Hardik.Harpal]

Please find modem trace NCS 1.5.1. More info (LTE Physical, MAC, RRC, NAS signaling) is required by JIO. Info already there in PCAP is not sufficient for them to analyze issue. 

 trace-2021-08-12T09-23-55.246Z.bin

trace-2021-08-12T09-23-55.246Z.pcapng