Virtual Safe: Unauthorized Walking Behavior Detection for Mobile Devices.

The prevalence and monetary value of mobile devices, coupled with their compact and, indeed, mobile nature, lead to frequent theft due to a lack of proper anti-theft mechanisms. Currently, there only exist damage control efforts such as remote wiping the device's memory or GPS tracking, but nothing to notify users of theft while it takes place. We propose such a mechanism which utilizes the unique walking patterns inherent to humans and differentiate our work from other walking behavior studies by using it as first-order authentication and developing matching methods fast enough to act as an actual anti-theft system. We test our system with the aid of 45 volunteers and demonstrate detection of unauthorized movement within 10 to 20 steps with an accuracy of 96.4 to 98.4 percent, while simultaneously distinguishing owners as themselves with 97.8 percent accuracy.