Clever hack, but a lot of people are misinterpreting what’s going on here. These devices emit some very low level of 125MHz energy during normal operation. This software is simply turning that on and off, but not doing anything to increase the amount of emissions.

Presumably the 125MHz emissions are within the FCC allowed envelope anyway, so this isn’t doing anything to exceed normal emissions limits. This only works in a quiet RF environment, as noted in the README.

There is no need to be concerned about this signal reaching aircraft or otherwise interfering with normal transmissions.

Very nice, could be used for exfiltration with some tuning.

The most advanced example of this kind of inadvertent transmission I've seen is Fabrice Bellards DVB-T transmitting with a standard VGA card:

https://bellard.org/dvbt/

(Note: As coderjames points out this could be dangerous tinkering. There is typically steady-state noise at 125MHz from Ethernet so it's not that we're putting more energy into the spectrum with this, but adding signal in the form of morse code could draw a lot of attention/distraction to pilots and ATC in the area.)

FWIW very brief example of 125MHz tone loss when going to 10MHz demonstrated here when my slow internet gets done uploading:

(Unpleasant sound warning)

https://youtu.be/JmyA5QEtAxA

Please don't try this at home! 118 MHz - 137 MHz is a protected Aviation band across the globe for airplanes to communicate with air traffic control. We already have enough industrial noise problems in this band; please don't contribute to pollution of protected spectrum. You will be interfering with the safe operation of the airspace.

See: https://en.wikipedia.org/wiki/Airband

Using an Ethernet line toner on a hot line in certain areas of Kuwait generates output from a local AM radio station.

As someone recently participated in an EMC measurement, I really don't understand how anyone passes these tests without some kind of cheating (using double-shielded, very expensive industrial cables + hacking with functional earthing).

In college we had an I2C to ethernet adapter on our drone testbed that caused all sorts of RF interference for us. We eventually wrapped the whole fuselage in a farraday cage so that the datalink and flight controls wouldn't be overwhelmed. It was responsible for transmitting data at a 1 Hz rate, and we could visualize the interference on a spectrometer over a broad range of Rf at exactly 1 Hz.

Anyway, we totally could have made a transmitter out of that thing.

Just out of curiosity: the tone appearing is me connecting my CE-certified Dell computer to a CE-certified Asus router, using a standard cat5-e cable.

https://ibb.co/0rxKq6L

(CubicSDR ran on the same DELL machine, however both tones disappeared after disconnecting the antenna from the SDR)

Of course, on the RPI there is the good ole GPIO4 abuse: https://tutorials-raspberrypi.com/build-raspberry-pi-radio-t... Works a lot better and has been around a long time.

This is a neat side channel attack for data exfiltration. The author is a radio amateur (Poland) and would do well to look at FT8 or other error correcting CW modulations other than simple Morse code. I would estimate you could pick up a signal at nearly a kM using such a scheme.

Would the cable length affect the tuning or is 125MHz specific to just the switching on the silicon?

While this may not be of practical use for most of us, spy agencies and the like would be interested in this or and other ways to exfiltrate data from air gapped networks that are not connected to the internet.

Previously posted here: https://news.ycombinator.com/item?id=25012469

The original Ethernet used similar hardware as ham radio. AlohaNet predates it of course, but Thicknet used local RF loops.

Billion dollar idea - RoI, Radio on Internet!