基于智能手表运动传感器的新型攻击及其防范

发布时间：2019-07-01 18:23

【摘要】：智能手表上搭载了加速度计、陀螺仪等运动传感器,且会随着手部运动而产生位移,使通过智能手表窃取用户外部键盘输入成为了可能。研究了智能设备上基于6轴加速度计/陀螺仪来推断用户键盘输入的可行性,建立了用户进行键盘输入时的运动模型来对手部运动进行分类识别。当用户执行敏感输入,如在数字键盘上输入密码时,可根据传感器信息推断出密码。经实验验证,该方法可有效识别4位PIN码,证明此安全问题值得人们重视。
[Abstract]:The smartwatch is equipped with accelerometers, gyroscopes and other motion sensors, and will shift with the movement of the hand, making it possible to steal the user's external keyboard input through the smartwatch. The feasibility of inferred user keyboard input based on 6-axis accelerometer / gyroscope on intelligent device is studied, and the motion model of user keyboard input is established to classify and recognize hand motion. When a user performs sensitive input, such as a password on a digital keyboard, the password can be inferred from the sensor information. The experimental results show that the method can effectively identify 4-bit PIN codes, which proves that this security problem is worthy of attention.
【作者单位】：北京航空航天大学电子信息工程学院;
【基金】：国家重点基础研究发展计划(“973”计划)基金资助项目(2012CB315905) 国家自然科学基金资助项目(61272501;61370190;61173154) 中央高校基本科研业务费专项基金资助项目(YWF-15-GJSYS-059)~~
【分类号】：TP368.33;TP309

【参考文献】