Slashdot Mirror
Researchers Turn Smartphone Vibration Motor Into Microphone To Spy On You (softpedia.com)
An anonymous reader writes from a report via Softpedia: Two researchers from the University of Illinois at Urbana-Champaign have come up with a method to turn smartphone vibration motors into makeshift microphones, capable of recording the sound around them. The attack relies on using the vibration motor's coil to record incoming sound waves, which are then transmitted to the attacker, who then uses a processing algorithm to enhance the signal by reconstructing high-frequency waves. This is needed because the vibra-motor can only pick up low-frequency sounds, up to 2 kHz. Their method doesn't yield perfect results (4 in 5 people can understand the sounds) and also needs physical access to the device, but it puts in place the theoretical details needed to carry out and refine such attacks in the future.
It turns out that in addition to having vibration motors, smartphones also have regular microphones.
Who'da thunk?
It requires much more than simple "physical access." They hardwired the vibration motor to an analog input.
"National Security is the chief cause of national insecurity." - Celine's First Law
...why not just install a microphone connected to the LINE IN instead of wiring the vibration motor to it as they have done?
Nice try Obama!!!
Apps should have to request access to vibration functions anyway. I'm sick of link-hijacking popunders that can vibrate my phone from through my browser.
"especially because there are far easier ways to spy on most people" You mean like the purpose built microphone on every smartphone?
No one can "Spy On You" using this method on any stock production phone. The vibration motor is connected to an *output* of the chip that drives it, not an *input*. Additionally, that output is likely to be digital rather than analog, so even its direction could be magically reversed, the likelihood of the chip being able to process whatever signal the motor would produce in response to ambient sounds would be just about zero. And if someone was modifying your phone in order to hear your conversations, there are *much* easier, faster, more reliable, less convoluted ways of doing it - like piggybacking on the microphone that's already there.
The ability to use a vibration motor as a microphone is a technical curiosity, but it's not at all surprising to anyone familiar with basic electrical and electronic concepts. The researchers' work is a nice proof-of-concept which may find useful application at some point. But really, the title of TFA, (and TFS), is solidly in the province of yellow journalism. There are more than enough *real* reasons to fear for our privacy - there's absolutely no need to further stoke that fire with false fears like those being promoted here.
'The Economy' is a giant Ponzi scheme whose most pitiable suckers are the youngest among us and the yet-unborn.
If you can physically manipulate the device, plant a proper microphone. If not, this is irrelevant, as there is no A/D input connected to that motor. The whole thing is an utterly worthless stunt by "researchers" greedy for attention but lacking in actual scientific skill. Why does this get reported here?
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
I would say that this gets reported is worse.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
Obviously, but in what way with what degree of success?
Sometimes there doesn't have to be but a workaround can be useful. As an example a few years ago I attended a presentation on Transmission Electron Microscopy on some structures in teeth related to tooth growth right down to the atomic level. Calcium atoms could not be imaged - data was missing from the signal due to lack of resolution. However comparing computer models of defocused images of different structures could be matched with the real images with not enough resolution to be focused led to being able to derive the structure without actually being able to properly image it.
That's an example of taking incomplete data and generating something that could have produced that incomplete data. If you can get a very large amount of the incomplete data, the missing signal is repetitive and you can model the degradation it's possible to model artificial data going through the same process and match it to the real data. I'd be interested to see what they are getting and how.
The important thing is that the low frequency noise is not always going to be the only clue and there can be some expectation as to what the original source is going to look like - eg. model of human speech and matching from running that through a filter matching the expected degradation. How the clues are put together could be interesting.
Yep. There's no way an operating system could possibly lie about that permission. You're totally secure as long as that checkbox is set to "off".
No sig today...
Another obvious 'easier way' is the speaker.... speakers can readily be used as microphones. I would like to hear the explanation as to why they would use the motor instead..... maybe because it sounds novel?
love is just extroverted narcissism