jcreedI did the math on my the -107dBm of signal my meshtastic device (heltec meshpocket: fun toy, recommend) was successfully receiving messages with, and that's 20 x 10^{-15} watts. That's not 20 milli-, not 20 micro-, not 20 nano-, not even 20 pico-, but 20 FEMTOwatts, y'all. That's absurd. One Joule in 50 trillion years. I hope I'm not overselling because I did the math wrong.
Joshua Wise@jcreed Jason I have wanted to give a !!Con-style talk on how insanely small GNSS signals are because I just cannot even begin to fathom it.
-107dBm is pretty good for a radio protocol carrying data -- I think Bluetooth LE tends to bottom out around -105 dBm, with most receivers being sensitive more to like -90 or -95 dBm. but once you only have to correlate *known* data, holy shit, you can get to the truly bonkers stuff.
if you have no idea where you are or what time it is, Sony's receivers can get a lock down to -149 dBm [1]. if you already have an almanac and vaguely know what time it is, you can get a lock down to -163 dBm. and once you already have a lock on some satellites and you're just tracking them, you can keep tracking them down to -167 dBm.
-167 dBm! what the FUCK! and they can do this while consuming just 6 mW of power!
IMO this is one of the most magical things we have ever built.
[1] https://www.sony-semicon.com/en/products/lsi-ic/gps.html
