QMI indications delivered at multiples of 30ms

[Lenzen, Léon]

Hello, 

while investigating some clustering in measured timings based on QMI indications, I noted that indications are delivered at multiples of ~30ms. This means that the time between some indications will be e.g. 680ms, 710ms, 740ms, 770ms, 800ms, 830ms, etc. depending on network parameters. 
I use my own GMainLoop and iterate is triggered every ~100µs (0.1ms). The clustering happens on EC25-E and RM520N-GL models from Quectel. 

Does anyone know where the clustering originates from? Is it some mechanism in libqmi or glib? Is it a specification of the QMI protocol? Is it hardware-related? 

Best regards
Léon
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3798 bytes
Desc: not available
URL: <https://lists.freedesktop.org/archives/libqmi-devel/attachments/20241010/a43a3413/attachment.bin>