[Cryptography] Radios vs. computers

[Henry Baker]

At 07:37 PM 5/26/2016, Bob Wilson wrote:
>>Thanks for the info, but this device isn't a *radio*, it's a *computer*,
>>and a really low power one at that.  So far as I know, there's no antenna
>>anywhere on this device.  I'm only interested in hacking its f'ing clock!
>>
>>If this device is "transmitting" anything at all, then there's something
>>dreadfully wrong with it, and/or someone else has already hacked it.
>
>There is no computer that is not a radio transmitter.  For a personal computer to be sold in the US, all that fancy metal stuff around the cabinet edges, the conductive layers put on a plastic case, etc., are all intended to try to control what it radiates so as to meet FCC specifications.  Your computer will have been certified to meet certain requirements, differing depending on the expected use environment.  Even a little "wall wart" power supply that you plug into the wall to charge a cell phone or to run a small appliance probably has enough high-transient switching going on that it is radiating fairly strong radio signals.  (Older linear supplies don't radiate much but they use more expensive copper and iron so they are disappearing.)
>
>I can remember hearing music played by the ORACLE at Oak Ridge in 1953, I think it was, by turning on a nearby radio and running appropriate code, and a decade later similarly for the little IBM 1620.  What you want to do may not depend on your computer transmitting, but someone else wanting to hear what you are doing might well take advantage of it.  Try putting a radio near your computer(s) and running a variety of programs.  The radio will definitely pick up signals, but whether it puts out sounds that you can hear depends on how it is choosing to demodulate those signals as well as what frequencies it detects.  AM will more likely produce a sound you can hear than FM would, but the signals are there even if they don't get demodulated in a way your ears pick them up.

(Yes, in my first job as an IBM 1401 nanny, we listened to the AM radio to see what the computer was doing.  Nowadays, you probably need a downconverter, but the idea is the same.)

All true, but other than regulating for RFI (interference), computers aren't legally radios, and don't fall under the same reqs.

I would imagine that many dimmable LED lights generate far more RFI than computers -- due to their use of pulse-width (or any other type of modulation) in order to get the dimming capability.

Everyone on this list should be afraid -- be very afraid -- of LED lights in their homes, as the hack to make them into audio bugs (which can be listened to from miles away with a decent telescope) is quite trivial (due to poor design, many LED lights might already be modulating the LED with sound frequencies purely by accident!).  Only slightly more difficult is converting an LED light into a wifi bug.

(If the intel agencies can already listen to audio waves by logging minute vibrations of window panes, then modulating an LED light is child's play.)

Such modulated LED light can still be detected and read even when window blinds are shut, in a manner similar to being able to detect which TV channel someone is watching based upon correlating the modulated light from someone's window with the overall brightness of a particular TV channel.