[HN Gopher] GhostTouch: Targeted attacks on touchscreens without...
       ___________________________________________________________________
        
       GhostTouch: Targeted attacks on touchscreens without physical touch
        
       Author : zdw
       Score  : 43 points
       Date   : 2022-06-02 22:41 UTC (2 days ago)
        
 (HTM) web link (www.usenix.org)
 (TXT) w3m dump (www.usenix.org)
        
       | xwdv wrote:
       | I can sometimes touch buttons on my iPhone screen without
       | actually touching them, at a distance of a few millimeters.
        
         | kurthr wrote:
         | Unless you're using a screen protector... something is quite
         | wrong. For large objects a barely acceptable false contact is
         | 0.5mm. There are conditions (moisture on the screen, cracked
         | electrodes, "floating" on a desk) where this may occasionally
         | happen, but if there is a specific location/area it's likely
         | caused by a fault.
         | 
         | Screen protectors (and their adhesives) can do pretty weird
         | stuff to the fields. Some designs may try to "calibrate" them
         | out. That can also cause problems. Similarly, if you use a
         | moisturizing hand lotion, that can leave a conductive film on
         | the surface, which is sensitive to relative humidity.
        
           | excalibur wrote:
           | > Unless you're using a screen protector...
           | 
           | Isn't everyone? I feel like the percentage of smartphone
           | users who don't usually have some sort of protection over
           | their screen is vanishingly small.
        
             | giraffe_lady wrote:
             | No I don't see the point and have never used them. In the
             | first half of the smartphone era phones would accumulate
             | microscratches but it took a couple years before they
             | became distracting to me and that was around when I
             | replaced phones anyway.
             | 
             | Since, idk, 2016-ish? the screens are resilient enough that
             | they don't scratch in a way that's noticeable while using
             | it, only if you're specifically looking for scratches. Sand
             | is the big exception, I basically keep it in a ziplock at
             | the beach. Otherwise I don't worry about it much.
             | 
             | I don't really think they meaningfully improve your chances
             | when you straight drop the phone. I try not to drop them.
        
             | kingcharles wrote:
             | I tear that shit right off. I hate anything between me and
             | the screen. The manufacturer didn't deem it necessary,
             | although they might have profit motives for that.
        
             | upbeat_general wrote:
             | If by vanishingly small you mean the majority of people who
             | own a phone them sure.
             | 
             | I'd guess the tech literate crowd is much more likely to
             | have a screen protector than a regular person.
        
       | somehnacct3757 wrote:
       | It's presented adversarialy, but perhaps there are practical
       | applications for this technique, even at the short distances.
        
         | hunterb123 wrote:
         | this is essentially already used for certain tablets with
         | styluses for a hovering over items when the stylus is near the
         | screen but not touching it.
         | 
         | not sure if it's the same process, but same effect.
        
         | excalibur wrote:
         | I wonder how effective this method would be at controlling
         | devices with cracked screens that don't respond well to actual
         | touch
        
       | megous wrote:
       | BTW, touchscreen can be another channel for communication (via
       | Goodix HotKnot).
       | 
       | https://www.youtube.com/watch?v=2ttPycepnho
        
       | unixpickle wrote:
       | > We can inject targeted taps continuously with a standard
       | deviation of as low as 14.6 x 19.2 pixels from the target area, a
       | delay of less than 0.5s and a distance of up to 40mm
       | 
       | Seems like 40mm isn't a very far distance. I understand that this
       | is a PoC, but it seems like making this work from a distance is
       | more "important" than having a low delay or super accurate
       | precision.
        
         | kurthr wrote:
         | This is because capacitive touchscreens are inherently a near-
         | field phenomena. They are designed to detect extremely small
         | capacitances (<fF) at relatively low frequencies (<500kHz)
         | relatively low voltage (<5V) at high report rates (>120Hz) in
         | the presence of very significant display noise (5-10V >1nF
         | coupling). They require >120dB of OutOfBand rejection to
         | operate. Cost per transceiver is ~$0.01.
         | 
         | They fundamentally measure the channel between arrays of
         | neighboring electrodes. That makes them self shielding (like a
         | Faraday shield) since they typically are designed to be
         | relatively immune to uniform changes in coupling. From any
         | distance comparable to the size of the touch screen the
         | differential coupling to electrodes falls off exponentially.
         | Near the screen they fall off 1/d then 1/d^n dominates farther
         | out.
         | 
         | In order to avoid interfering with other parts of the device,
         | they tend to be highly encoded narrowband signals and change
         | (to known good) frequencies when they detect interference. That
         | makes most narrow band techniques less effective. High voltage
         | impulsive noise is most likely to have an effect, but
         | triggering ESD detection/rejection might create a sweet spot in
         | amplitude.
        
         | dwheeler wrote:
         | At 40 mm, it seems basically useless as an attack in almost all
         | circumstances.
         | 
         | However, I wouldn't write it off completely, because if
         | somebody else can extend its distance further then I can see it
         | becoming a concerning attack.
        
           | kurthr wrote:
           | I'd agree, requiring in-room access makes it not that big a
           | deal. There are other attacks (like measuring the radiated
           | emissions) to detect finger coupling remotely that I'd be
           | more concerned about. Those could detect private user input
           | and could be hidden nearby.
        
         | nine_k wrote:
         | If it's enough distance to penetrate your pocket (which I'd say
         | more like 3 mm), it may have interesting implications.
        
           | megous wrote:
           | Touchscreen should be off in your pocket.
        
             | bsnal wrote:
             | All phones I've used in the last years turn on when you tap
             | the touchscreen once or twice while it's off.
        
               | bikingbismuth wrote:
               | (Hopefully)Most phones are locked when they are woken up.
        
             | madacol wrote:
             | Call the target, screen turns on, ghosttouch to answer
             | before it even rings
        
       ___________________________________________________________________
       (page generated 2022-06-04 23:01 UTC)