@philip max yeah this is a solid question and one I've thought about too when doing link characterization work.
Short answer: yes, you do need to be careful about interpreting RSSI directly on the MultiNav Pro+ because the external LNA in the RF front-end shifts your effective noise floor significantly compared to a vanilla STM32WB07 setup. What that means practically is your RSSI numbers will look "better" than what a commodity module would report at the same distance — the LNA is pulling in signals the bare chip would have buried in noise, so the dynamic range you're working with is wider. That's great for actual link performance obviously, but if you're trying to build a propagation map you want to make sure you're calibrating your reference point at a known short distance first rather than trying to use textbook free-space path loss assumptions against the raw RSSI numbers.
For the logging side, the firmware repo on GitHub is a good starting point — if you're building your own logging layer on top, I'd suggest sampling RSSI repeatedly over a short window at each position and averaging rather than taking single-shot readings, because the front-end amplification also means you're slightly more sensitive to multipath fluctuation than you might expect. A rolling average over even 10-20 samples smooths that out enough to get useful spatial data.
One more thing worth doing: take your baseline calibration reading in both directions (node A → node B and B → A) since asymmetric gain configurations can produce slightly different reported values depending on which end is doing the measuring.
