I am fascinated by techniques to indirectly capture data from computers. In college we had software to play sounds on a radio placed near a processor or printer. The signals from the device generated radio-frequency data one could hear on a radio. Clever programmers even made music from those sounds. Some people still do this for fun.
My interest was really piqued when I learned about Van Eck phreaking in the 1980s. The idea that Wim Van Eck came up with was simple – computer displays (Cathode Ray Tubes or CRTs in those days) emit electromagnetic signals that could be captured over the air from a distance and displayed on a television. His paper linked above is fascinating. In those days there were even ads in consumer electronic hobbyist publications for tools to do the phreaking.
It also turns out that LCD displays do indeed generate recoverable signals, making them susceptible to eavesdropping, too.
We talk about some interesting methods for capturing user input (e.g. keystrokes) in Learning Tree’s System and Network Security Introduction. Some examples include looking at signals transferred to power lines, examining the vibrations of a laptop to see what keys are pressed, among others.
Imagine my interest, then, when a Wired article appeared entitled Stealing Data From Computers Using Heat. I envisioned looking at keyboard heatmaps or some other real-time information capture. While that wasn’t what the article was about, it is fascinating nonetheless. It turns out some researchers found a way to cause computers to generate predictable amounts of heat and to have that heat detected by thermal sensors on another (very) nearby computer. The maximum information transfer is a slow eight bits per hour, but that could easily be enough in some circumstances.
Consider, as the Wired piece notes, two computers side-by-side. One is on a computer network and one is isolated from all networks, a condition sometimes called an air gap. This situation is used in secure environments where work done on the isolated computer is so sensitive it must be kept only on that computer. A similar situation exists where the two computers are on different networks: a very secure one and a less secure one. I spoke with a colleague who managed just such situations.
If both computers were compromised and the isolated or very secure one could be used to transfer data – for example a password – to the other computer, traditional security methods could be easily defeated, albeit slowly. I doubt that this is practical, and the distance between the computers would have to be very small. But, the proof-of-concept is at least interesting and may lead to some kind thermal shielding of secure computers.
The article didn’t end up being what I expected, but it was still interesting. I continue to be fascinated by covert eavesdropping. If you are interested in such things, and if you have any interesting examples or stories to share, we’d love to hear them in the comments below.
To your safe computing,