基于Android权限机制的动态隐私保护模型

针对Android平台自身粗粒度权限机制的缺陷以及缺乏有效预防程序间隐私泄露机制的问题,提出一种改进的细粒度权限配置机制与隐私数据动态着色隔离相结合的Android隐私保护模型。通过对Android应用程序权限进行细粒度的动态配置,阻断隐私数据从程序内部泄露的途径,利用隐私数据着色跟踪实现对程序间传播的包含不同隐私权限标签的消息的隔离控制。通过大量实验的反复测试,该模型可以有效保护Android程序内部的隐私数据,及时发现程序间权限提升攻击进而实现隐私数据隔离,从而全方位实现Android隐私数据的保护,并为以后相关研究提供了新的方向。

Dynamic privacy protection model based on Android permissions

The privacy mode in Android has some defects and is lack of effective mechanisms to prevent loss of privacy between applications. This paper proposed a new privacy protection model by using the fine-grained permissions and privacy data tainting. Firstly the method could prevent loss of privacy from a single application by dynamically configuring the permissions in a fine-grained manner. By leveraging the track of the tainted privacy data it could prohibit the propagation of the privacy data between applications which had different permissions on the accordingly privacy tag. Repeatly experimental tests show the model is able to effectively protect the privacy of data within the Android application, and to detect the privilege escalation attacks between applications so as to isolate the private data. As a result, the full range of private data within Android is in protection, and leads a new direction for related research in the future.