Research On Android Malicious Application Detection Based On Feature Function

With the "Internet plus" era, the mobile Internet ushered in a new development opportunity. Smart phone as one of the main entrance of the mobile Internet is developing rapidly, and its equipment performance and penetration are greatly improved. As one of the most popular mobile operating system in the Android system, bringing convenience to people at the same time, is also facing serious security threats. Android system security is a problem that can not be ignored and urgently needed to be solved in the field of information security. As the main component of Android system security, malicious application detection has great research significance. The main research contents of this paper are as follows:Firstly, this paper briefly analyzes the Android system, focusing on the analysis of the security mechanism of Android system. Although the security mechanism of Android is more complete, but the need for flexibility and other aspects of the system can not be very good protection. The situation of Android system security is very grim.Secondly, a malicious application detection scheme based on feature function is proposed. Through the research of Android malicious application, the discovery of malicious and non malicious applications in the call to some system functions, there are a lot of differences. The call of system function can be used as the basis for classification,and can be used to determine the category of the application.Then, the design and implementation of malicious application detection scheme based on feature function is completed. Through the analysis of the function of Android system,the characteristic function table is established. Through the use of the paper design of the automatic decompile tool access to the source code, and Sunday algorithm for pattern matching, identification of system function call, and calling the KNN algorithm to classify the application operation,completed the experiment.Lastly, the proposed scheme is verified by experiments. The experiment mainly completed the K is worth to select, as well as the determination of the experimental results.Through the experimental results, we can see that malicious application detection scheme based on feature function can effectively classify Android applications, and when the Kvalue is 7, the detection accuracy is as high as 90.8%.