Researchers from Alibaba Security demonstrated at Black Hat 2017 a sonic hacking tool (via Ars Technica) that could attack your iPhone or Android device, as well as other devices. Products with microelectricalmechanical (MEMS) sensors like gyroscopes and accelerometers are affected, which means that anything from drones to air bag sensors in vehicles are fair game.
Sonic Gun
The sonic gun works by tuning an audio signal to the resonant frequency of the MEMS sensor, which can disable or manipulate the sensors into outputting false data. This inevitably causes the devices and software to fail.
Wang Zhengbo and Wang Kang from the research team demonstrated the tool onstage against the iPhone 7, Samsung Galaxy S7, VR headsets, a DJI drone, a Xiaomi hoverboard, and a toy robot. The team notes that the best long-term solution to their gun is the development of new MEMS sensors that could be resistant to sonic attacks. You can see the device in action in several videos.
The video below was included with the researcher’s paper, and shows their sonic gun interfering with an iPhone 7.
Sonic Gun vs. Robot Toy
Sonic Gun vs. L.I.U. MI Hoverboard
Because of the hoverboard’s thick plastic shell, the researchers had to physically press the sonic tool against it. That makes it unlikely the tool could be used as a long-range weapon in its current state. There’s little doubt, however, that criminal and state actors will try to build off this research.
@Lee and @MacFrogger:
A principal reason for conducting such research, particularly for mission critical devices and systems, is the recognition that real life is far more complex and creative at throwing things at you that isolated imagination would never have anticipated. One can either be reactive, and address problems piecemeal as they crop up and forever play catch up to the real world (expensive and sometimes fatal) or keep throwing everything one can at the thing, particular things that it is likely to confront, until you break it.
Once done, you can either let the breakpoint define your operating parameters (routinely done especially if the fix is not cost effective or timely) or, especially if it represents a potential threat that real world use cannot avoid, fix the limitation or develop a new piece of kit or next gen upgrade.
Bottom line: better to know, before you deploy on mission, what might kill you than to discover it in the field with no options.
Could be MacFrogger, but the developers wou;d probably be under an NDA. There are such things out there
https://www.wired.com/2007/04/navy_patents_ca/
Lee: Could be the military funded it in the first place; e.g. to knock a hostile drone out of the sky.
Then why do the reasearch? Looking for a way to counter it I suppose.
There’s definitely a fine line that ethical hackers need to walk before they’re considered one of the bad guys. I think it’s mainly that the good guys research stuff like this in order to help prevent it.