Research On Malicious Code Similarity Matching Calculation And Power App Security Test System

In recent years,mobile terminals and applications in the power grid company's various business systems have been promoted continuously.Malicious code programmers use the Android open platform to develop a lot of malicious codes,and then interfere with the user's mobile phone illegally.Once the phone terminal is infected by malicious codes,attackers("hackers")can maliciously have access to large number of user privacy information(including user account password information,the user phone number information,etc.)by the way of illegality[1],even to block text messages,to delete the user's mobile terminal applications,which lead to the occurrence of a series of serious incidents of harm.Therefore,the research on the Android malicious code detection technology and the development of power grid for the provincial power grid App security test system have some practical significance.In this article.first of all,I get the bytecode file through the tool of apktoolkits,then,from the file of smail.I extract API instructions and achieve the feature extraction of app's application program interface(API).what's more,the configuration file of AndroidManifest.xml needs to be extracted,and I get the authority of application when the installation is running.Secondly,according to the extracted feature,I make a feature information table and use the rough set theory(Rough Sets Theory,RS)on the App feature weight calculation,then,with the help of euclidean distance,I calculate the similarity between the current characteristic case and the characteristic attribute in the case library;The random forests algorithm(RF)is used to match the application characteristics.A certain number of samples are extracted from the original training set and are trained as root nodes.They are continuously trained until all nodes are traversed or at the end of the training,so as to achieve the characteristics of leaf nodes and case libraries in the characteristics of matching.I use an intelligent App to extract features and conduct the simulation of RS and RF algorithm,both of them get the ideal effect in the Android malicious code detection.And the latter is more accurate.Finally,at the same time,Conducting the comparative test between the automation test platform and Ali poly security platform by enterprise credit app,mobile cloud app,WeChat app.It can test loophole of malicious code features.The experiment shows that the developed testing platform achieves the expected goal.