Is it possible to add LoRaWAN FUOTA for STM32L0 (B-L072-LRWAN1)?

[Deloins]

I recently looked into adding FUOTA to a LoRa project I'm working on. Right now, we are using the B-L072-LRWAN1 discovery board. I-CUBE-LRWAN only gives example code for the STM32L476. Is it even possible to do this on the STM32L072 embedded on the discovery board given the constraints in Flash Memory and RAM (192kb and 20kb respectively). Please let me know.

[DPard.1]

Hi Deloins,

​

I am also interested in this question, any answer by ST?

Thanks.

[Tesla DeLorean]

192KB is quite a lot of space, how large is your firmware image currently?

Most FOTA methods need to be able to stage an updated copy, either in internal flash space or on a SPI NOR Flash type device.

[Deloins]

Thanks for you response. Our firmware image uses 104 kb in internal flash and 10kb in RAM. Judging by application note AN5411 it seems that you'd need quite a bit more space to accomodate the updated firmware as you would basically need twice the size of the original firmware image to store a copy. See the attached image. Besides the STM32L0 doesn't have an FSMC interface so adding an external flash chip wouldn't work if I am correct. Please tell me if I'm wrong cause I could be mistaken. Maybe stripping away a lot of features would work but then it might as well be easier to upgrade to the STM32L4 right?

[Tesla DeLorean]

A SPI Flash would only need an available SPI interface, and a chip-select for sharing the resource, rather than a full F(S)MC

The STM32WL might be a better choice, I've heard of at least one implementation using a similar foot-print to the Murata for the prototype.