Using AI and KiwiSDR TDoA

[Andrew (grayhat)]

Maybe it's a crazy idea, but... let's start from this

https://www.rtl-sdr.com/kiwisdr-tdoa-di ... ublic-use/

now, let's suppose to put together some code which, by leveraging the above and some map API, may feed an AI to gather further infos (e.g. using search engines and publicly available data) to improve the operation of finding the source for a given signal in short (if not real) time

thoughts ?

[13dka]

Sounds intriguing. My personal experience with the multilateration option isn't great though, there are many practical problems (such as receiver availability and their often less than optimal QRM situation, quality of GPS reception and whatnot) that seems to render this method very unreliable. I once tried to mulitlaterate the known position of the Rohrbach shortwave transmitter under what I deemed optimal conditions, with receivers surrounding the location close enough to receive the groundwave signal. Still, every single of my attempts was at least 100km off and none of them hit the bullseye. When things went only a little wrong, I couldn't have told for sure if the transmitter is in Germany at all.

I also tried to replicate the measurements that YouTube guy made on the allegedly Kuban 40m-band jamming and had a hard time finding KiwiSDRs in what I deem apt positions and ended up with enough ambiguity that I wouldn't be sure whether to blame the Cubans or some operation at the US gulf coast. Of course, maybe I just did it wrong because setting that up in a meaningful way isn't exactly trivial either.