Strange behaviour caused by ADR activation

Hi there!
I’m having and strange behaviour of my node when it joins to the network and Adr is set. I’m using the MCCI LoRaWAN LMIC Library, and every time that the node joins to the network, it sends several(around 8) unconfirmed uplinks, and I can´t find the reason.

In addition, every unconfirmed uplink sent, receive an Unconfirmed downlink from the network server, with the field nbRep: 1.

I don´t know if this amount of unconfirmed uplinks is defined in the library (I couldn’t find something related to) or the node is just sending all those unconfirmed uplinks because the gateways ask for them.

I’m working with AU915 frequencies , and dataRate is defined between 0 and 5 on the service-profile

I’d really appreciate any hint!
Thanks in advance.

Best regards!

My assumption is that during these uplinks and downlinks, the NS is tuning the data-rate and tx-power :slight_smile:

Anything sent on fport 0 is network level housekeeping traffic. What you are seeing is definitely not optimal behavior (the design intent of LoRaWAN is that ADR “tuning” happens in a lazy fashion over a substantial period of time riding on useful traffic, not aggressively at the start) but to start to understand what is going on you need to expand those messages and see what they contain in the way of content.

If I had to guess, either the exchanged of MAC commands/responses is broken, or it’s a correct exchange but one end (probably the node) is responding immediately with a housekeeping only packet, when what it should be doing is only attaching that response to the next packet sent with useful application data. Vs what’s happening right now when it seems to be uplinking again immediately after each downlink.

Are you using an up to date version of the MCCI LMiC? Very old version of the LMiC code had some “interesting” bugs that have since been fixed.