I have an RFM95 based device using the MCCI LMIC v4.0 library. The device is periodically switched on/off and the join requests are not stable. The device has to send up to 5 join requests for the connection to be established.
One clue is visible in the output:
DevNonce already used
For the others it isn’t clear from your output, however it is probably timing related.
BTW, your requests are spaced too close together, a gateway has to adhere to the same duty-cycle a node is subject to. So some of the responses can probably not be transmitted due duty cycle limitations.
Are you node and gateway close? Make sure there is at least 5 meters and preferably a wall (or two) between them to prevent overload of the receiving circuits.
You should not join every time when the device power cycled, you should save the status and restore them when you power on. By joining too much you may be blocked from joining by the gateway. And also is a waste of air time.
imho the gateway is not able to block Join Requests. A few days ago my gateway received more than 1300 Join Requests from one node within abt. 4 hours.
The gateway just receives the Join Request and sends it to the TTS CE server. Blocking can only be done by the server.
Wow, that’s about once every 11 seconds.
Some gateway can block OTAA request (also uplink) (at least the one our company deploys, we can set a minimum interval for an OTAA request being forwarded). We use that to save data (since we are using cellular network for the gateway) and prevent rogue node from spamming.
I just don’t understand why there are always nodes that spamming OTAA and SF12 uplink with like 50+ bytes payload…
1 Like
Indeed. I didn’t believe what I saw when looking into the log-file of my gateway.