基于程序分析的Android应用恶意行为检测

随着Android智能手机的普及，手机应用的安全隐患也日益凸显。提出了一种基于程序分析的Android应用程序检测方法，用于检测Android应用程序中的恶意行为。通过预处理剔除不存在恶意行为的应用程序。对通过预处理阶段的应用程序，模拟执行应用程序中的字节码指令，构建出函数的摘要信息，最终在构建的函数摘要上使污点传播算法，检测应用程序中的恶意行为。实验结果表明，该方法可以有效检测出Android应用程序的恶意行为，具有较高的实用性。

Malicious Behavior Detection of Android Application based on Program Analysis

With the popularity of Android smartphone, the application security of smartphone becomes increasingly prominent. The paper presents a program analysis-based method for detecting malicious behaviors in an Android program. The applications with no malicious behavior are filtered out by pre-processing. As for the applications that have passed the filtration, the bytecode instruction is executed by simulation, thus to construct function summary. Finally the taint propagation algorithm is used to detect the malicious behaviors in the applications. The experiments indicate that this method is practicable, and can effectively detect the malicious behavior in Android applications.