Design And Implementation Of Security Detection System For Android APP Based On Network Behavior

With the rapid development of mobile technology, smart phone has becomecommon, which Android becomes the most popular operating system. Users coulddownload and install the third-party applications to meet their functional requirements.However, since the third-party platform lack of supervision and regulatory on theapplication, Android malware increase dramatically. Among them, a large proportionof malware achieve malicious purposes through transferring the malicious data andattack files with network behavior, which has a serious threat to the privacy of usersand property.This paper designs and implements the security detection system for software onAndroid based on the network behavior. Malware bring about attacks through theAPIs provided by Android, which is hided in the normal behavior sequences. Againstto this problem, this system is achieved by extracting and analyzing the networkbehavior of the applications in the running time. Firstly, we should determine theAPIs which are need to be tracked through analyzing the methods and purposes ofmalware attacks at this stage. Then, we will set up Hooks by the log mechanism totrack and extract the behavior sequences of targeted APIs, which are invoked by theapplication during the running time. Finally，we will use the machine learningapproach to classify these behavior sequences to ensure the security of the applicationand provide security strategy.This detection system has played a very good effect on detecting malware whichuse the network behavior to achieve malicious attacks. It also has the ability tomonitor and deal with malicious software and interrupt the behavior of an attacker,which could provider good protection for the user’s mobile phone security. Theexperimental results show that the system has higher detection efficiency anddetection accuracy.