Research On Analysis Method Of Android Application Permission

Mobile devices based on Android system are favored by more and more people.Meanwhile,with the rapid growth of applications such as chat and communication,audio and video,reading and learning,users' dependence on mobile devices is deepening,and more and more confidential information is stored in mobile devices.In order to analyze user needs or achieve their own purposes,some applications' developers access user data or system resources without notifying the user.The research on the first line of defense of Android security model,the permission mechanism,has become an important topic to protect the sensitive information security of users in recent years.In this paper,we analyze in detail the permission change of Android system,Android security mechanism,vulnerability of application permission and application security analysis based on permission.Two detection schemes for the vulnerability of Android application permissions are designed,and a comprehensive Android application risk assessment method based on permissions is proposed.The main work in this paper are listed as follows:(1)A risk assessment scheme of Android application based on permission was designed.The scheme extracts the system permissions declared by the application,the permissions obtained through static analysis and custom permissions by reverse-engineering analysis of the application.At the same time,the permissions used by executing application were extracted through dynamic detection.Then,this scheme performs quantitative risk assessment of applications from three aspects: permission combination with malicious intent,"over-privilege" problem and custom permission vulnerability.And the analytic hierarchy process evaluation model was used to calculate the weights of the above three aspects for assessment risk value of the application.(2)A permission promotion detection scheme for multiple applications based on permissions and components is designed.By decompiling and analyzing the application program,the permission declared by the application program,the component and the call information of the application program interface are extracted,and the component function call graph of the application program is constructed with the method of deep traversal to detect the path of permission promotion of the application program,which can be used to detect the possibility of permission promotion in multiple applications.(3)A comprehensive Android application risk assessment method based on permission is proposed in this paper.The system uses Python and Monkey automated test technology to carry out automatic risk assessment of the application program,and constructs data which sets with malicious tendencies such as composite permission data set,application custom permission data set and application component data set.This approach quantitatively evaluates the risk value of Android applications from the existence of application permission vulnerabilities and the possibility of participating in multiple application permission promotion.Experimental results show that the proposed method,compared with Androguard,can more accurately assess the risk value of application software.