4 ms·
Wi-Vi: See Through Walls with Wi-Fi Signals
- stephengillie 13y agoPrevious discussion: https://news.ycombinator.com/item?id=5824286 https://news.ycombinator.com/item?id=5824286 https://news.ycombinator.com/item?id=4334264 https://news.ycombinator.com/item?id=4334264 I'm still waiting for someone to make an app that does this -- would it require linking 2 phones?
- mirashii 13y agoFrom what I recall when I skimmed the original paper, you would need more general hardware than the wifi chipset in a phone. Actually, that this says it uses wifi at all is a bit misleading. It happens to work on the same frequency bands that wifi uses, but their experiments did not actually use wifi hardware at all and actually used USRPs. Getting off the shelf wifi hardware to do the same would require firmware and drivers that allow raw access to the radio signals, which no cards I know of come even close to. Additionally for their experiments they needed at least five antenna, so even if it were going to happen with phones you'd probably need more than just two.
- stephengillie 13y agoSo it's like how Xbee radios use the Wifi radio frequency range, but use their own protocols and are actually 1-2 MHz off from Wifi? Xbee hardware can be co-opted into Wifi transmission by an Arduino or more powerful processor: https://github.com/cjbearman/xbee-wifi-spi-arduino https://github.com/cjbearman/xbee-wifi-spi-arduino I wonder if Xbee devices would be more suitable to this task.
- pudquick 13y agoActually that driver is specifically for 802.11 (wifi) XBee modules like these: http://www.digikey.com/product-highlights/us/en/digi-international-xbee-wi-fi/108 http://www.digikey.com/product-highlights/us/en/digi-interna... It does not cause 2.4Ghz 802.15.4 XBee modules to act like a normal wifi (802.11) chipset. Just an FYI.
- throwawaykf02 13y agoActually, this is similar work but by different people. The links you posted are about work at University of Washington, whereas the OP is work done at MIT.
- dobbsbob 13y agoPut a SICP tent in your room (:
- zalzane 13y agoThis looks like it would make an excellent spying tool, if there isn't already a similar device available.
- zw123456 13y agoGreat, now the NSA not only can snoop on my emails but use my wifi hotspot to watch me too.
- Spooky23 13y agoGiven that secure buildings typically lack windows or have RF shielding on the windows, I'm sure this sort of capability has been around for a long time. Example: https://www.astic.com/sdfilms.php https://www.astic.com/sdfilms.php
- qrybam 13y agoI can see this being used for security. Imagine replacing passive IR motion detectors in the home with this!
- jffry 13y agoThe only advantage I can think of for this is that it could be concealed in the wall, where a passive IR motion detector requires a clear field-of-view of a room. The downside is that this is actively emitting microwave radiation, so it would be easy to detect, while a passive IR sensor would not be emitting any discernible radiation (assuming its internal electronics are properly shielded)
- nether 13y agoseems like something the CIA science directorate could have come up with, but never released to the public
- kyle_martin1 13y agoRF and EE here. This is nothing new. Or at least a motivated junior/senior in college would stumble upon this simple application of wave propagation theory pretty early on.
- jffry 13y agoThis is notable because it's an actual implemented system. Yes it works in theory, on the exact same principle that a radar speed gun works, but these researchers have shown that it is possible to implement in the 2.4GHz band. That's the first step to making it smaller and cheaper.
- kyle_martin1 13y agoIf you think this hasn't been cheaply implemented before then you're behind research in wave propagation brother. Search on IEEE if you're interested.
- nickpinkston 13y agoYou also get a grade A in snark! ;-)
- throwawaykf02 13y agoI know there has been tons of see-thru-the-walls work done over many years... But that used UWB and required specialized equipment and antennas, IIRC, whereas this uses "commodity" USRPs. Could you provide examples of the type of work you're talking about?
- mrcactu5 13y agowhy can it detect humans and not other moving objects? knowing how many people on the other side of the wall will change our sense of privacy
- jffry 13y agoThe linked article and video never said that it wouldn't work on non-human moving objects, just that it would work on moving humans (as it is a wireless relative-motion detector) It is possible that this tech is especially effective at tracking humans due to the fact that we're bags of mostly water and the specific frequency used interacts with water more than other nearby frequencies. (I'm not a microwave engineer, so I don't know this for sure)
- kyle_martin1 13y agoYou are correct. We can determine what an object is made out of or at least behaves like based on the reflected EM waves. Every piece of matter has an associated electric permittivity and permeability. These can, and often do, change with temperature and the other environmental variables.
- stephengillie 13y agohttps://news.ycombinator.com/item?id=5826632 https://news.ycombinator.com/item?id=5826632
- femto 13y agoCombating reflections (hence multipath) from moving objects is precisely the reason Wireless LANs use orthogonal frequency division multiplexing (implemented with an FFT). The environment leaves its imprint on the received signal and depending on your point of view: communications or radar, that imprint is either information to be retrieved (radar) or eliminated (communications). If you look at one of the early WLAN papers [1] (I'm an author), Figure 1(b) shows a frequency selective fade, meaning there is a dip, or null, in the received signal and some of the frequencies are missing. The way we created that null was by a person walking through the room. With a change of emphasis in the signal processing, a WLAN is a very good person detector. [1] http://wireless.ece.ufl.edu/twong/Preprints/00566198.pdf http://wireless.ece.ufl.edu/twong/Preprints/00566198.pdf