This is a 'tour de force' study for sure, but has a very limited scope (and reliability)- First, it only works if the attacker knows both the route(s) ahead of

This is a 'tour de force' study for sure, but has a very limited scope (and reliability)- First, it only works if the attacker knows both the route(s) ahead of time and the power consumption profile of the routes, which requires careful mapping of the region with a recording device. Second, its accuracy degrades drastically depending on the number of apps running -- they only tested with background apps, which already rendered the method only slightly better than a random guess, with an arbitrary app running in the foreground the power consumption goes bananas and so does their method. Storm in a glass of water, if you ask me.. (But you wouldn't know this by reading that abstract alone ;)) -@r2r