Study On Android Intelligent Mobile Terminal Application Security Judgment Scheme Based-on Behavior Analysis

With the rapid adoption of Android mobile intelligent terminal, the security issues have become increasingly prominent. More and more malicious Android applications are downloaded via various channels, installed into smart mobile devices. The malicious application is a serious threat to the security of smart mobile devices. Judgment of Android application itself has become an important safeguard of ensure Android smart mobile terminal security.This paper studies the security mechanism of the Android platform, the overall architecture and the Android platform, and then analysis the security threats that the Android platform is facing. For existing risk, the most effective way is to judge the security of the Android applications which are circulating in the major channels.This paper also analyzes the Android platform specific application behavior detection method. Currently, analytical techniques for Android smart mobile terminal applications are based on simple static code analysis or simple dynamic log analysis. However, a simple way to analysis Android applications have many problems:simple static code analysis always relies on existing malicious code sample eigenvalues or semantic logic, for unknown malicious applications anticipation is almost zero; because of the limitations of resources and time, simply rely on dynamic log analysis can’t fully analyze the application of pre-judgment.This paper uses static way and dynamic way to analysis Android smart mobile terminal applications security based on behavior. This paper introduces the SVM and KNN machine learning algorithms, respectively, from the perspective of the static and dynamic analysis Android applications. This scheme not only comprehensive analysis through the application code and semantic analysis, but also can detect unknown malicious applications valid identification. This scheme could ensure the safety of Android mobile intelligent terminal from the view of the security of application itself. Finally through the experimental method the paper validate the accuracy of the scheme. The results show that the accuracy of the scheme for the identification of unknown application is higher than the scheme which purely based on static analysis or dynamic analysis.