PowerSpy: Location Tracking Using Mobile Device Power Analysis

I've only scanned this on my phone, not really dug in to the math, but this strikes me as significantly more serious than the current comments here suggest.

Low power tracking is a significant part of the motivation behind mesh networks in other applications, and the authors here are quite familiar from my school days. I may have more to say in a few hours, but this is likely a Big Deal.

"We also assume that enough communication is occurring for the signal strength to have an effect on power consumption. This is a reasonable assumption, since background synchronization of data happens frequently in smartphone devices. Moreover, the driver might be using navigation software or streaming music. However, at this stage, it is difficult to determine how inconsistent phone usage across different rides will affect our attacks."

I don't find this assumption to be reasonable at all. There are plenty of days where my cellular data usage doesn't exceed 1MB because I'm either at home, at the office or moving between them. Most background synchronization happens while I'm using WiFi (and stationary). If your location tracking activity relies on me streaming music while on route, you would probably be better of waiting outside my apartment and following me around.

I appreciate the fact that other users have different mobile habits, but the authors didn't convince me that their scenario poses a feasible and significant threat to privacy.

Is the app actually available? It is said to be available on the Play Store after "unblinding" (I assume that means something like after publishing, it doesn't seem to be an actual word[1]) but presently I cannot find it.

[1] http://www.merriam-webster.com/dictionary/unblinding

I think this is a really long article for a relatively very small hack. For the most part location cannot be obtained, at max i think the victims route maybe identified with respect to the cell tower its connected to.

Also published by one of the authors: http://crypto.stanford.edu/gyrophone/