LTE, analysis of connectivity

  • This information might be of some value to some members of the community. I have tested my setup through the night.

    My code is pretty straight forward and based on @micksta 's example for an LTE Cat M1 setup in Melbourne (which is also where I am).

    I use Telstra's network, band 28.

    lte.attach() seems to hang sometimes, so I have introduced the WDT to take care of a connection that is taking longer than expected (or that will never be made...).

    I use the deep sleep mode for 30 seconds after doing all the required operations (mostly setting up the LTE connection). From the data below, we can safely assume that these take about 35 seconds (65s - 30s deep sleep). I could probably reduce some delays between AT commands in the code, but would decrease the reliability of the system.

    The behaviour is generally predictable, when the connection is made, then all is good (it will take pretty much the same amount of time to do all the data exchange). But sometimes the connection is not established and the WDT takes over. Several failures have happened in a row, multiple times.

    Overall, 19 operations out of the 365 took longer than expected for this test. So we are looking at a reliability of 94.79 % .

    Here is the data:

Pycom on Twitter