4 ms·
Yes unfortunately "model craft" is not defined anywhere in Part 97 that I can find. I can't say I understand the rest of the first paragraph - Part 97 doesn't
by threemux 3y ago
Yes unfortunately "model craft" is not defined anywhere in Part 97 that I can find.
I can't say I understand the rest of the first paragraph - Part 97 doesn't actually ban encryption, just "codes or ciphers intended to obscure the meaning of the communication", from which "telecommand of model craft" is specifically exempted.
I'm not as familiar with the Part 15 rules - do they allow high gain antennas? Since this case is about stationary command and control, I wonder if lower power + high gain wifi antennas at both ends would close the link.
- JoshTriplett 3y agoI'm saying that "control signals" and "telemetry" also aren't clearly defined, and while dedicated protocols like those used to control an RC airplane are clearly "control signals", it's not obvious if SSH-over-IP-over-wifi to poke around and debug an avionics system counts as either "control signals" or "telemetry". You could make a case that it is, but it seems much closer to the line and not obviously permitted. > Since this case is about stationary command and control It's not; this was the same system used while the rocket was in motion, and part of the point of using the same system was to make sure it's working before launching. > I wonder if lower power + high gain wifi antennas at both ends would close the link We were using high-gain antennas already, and IIRC that didn't suffice within the 100mW power limit.
- fragmede 3y agoI've always wanted to make a ssh-telnet hybrid that sends data plaintext but with an authentication mechanism so that commands get sent in the clear, to meet the requirement that it's not encrypted, while the authentication layer prevents others from spoofing commands. thinking about it, presharing keys and then using gpg to authenticate the message should work. just need client and server programs to make it convenient.