Design And Implementation Of Malware Detection System For Mobile Devices

Because of the open source feature of Android system,the third-party software market lacks supervision and unified standards which pose a serious threat to users' privacy and information security.Although mobile terminal manufacturers and malware detection agencies have their own detection methods,the existing technology cannot share malware detection information of different manufacturers in real time.Blockchain technology has the characteristics of high transparency and decentralization,which ensures that relevant data cannot be maliciously modified.Using this technology,real-time sharing of malicious software information can be achieved,which is particularly necessary for protecting the privacy and property security of users on mobile platform.In this thesis,the malicious software detection technology of mobile terminal is deeply studied.The main research work is as follows:(1)Aiming at the problem that the existing detection technology can not realize data sharing of malicious software,we combine Blockchain technology with malicious code detection technology,and propose a malicious software detection framework based on Blockchain for mobile terminal,which can realize rapid detection of malicious code in Java layer of mobile terminal.The method of feature representation and extraction of malicious software is given,and the multifeature model of malicious software family is established;The multi-feature detection algorithm of malicious code in Java layer is proposed,which can detect and classify malicious software family.The experimental results show that the algorithm has good effect in detection efficiency and precision;The Blockchain structure and the corresponding workflow of digital signature and verification and consensus mechanism are designed to ensure data security and untouchable modification.(2)Aiming at the problem of difficult detection of low-level code in mobile terminal,a framework of malicious software detection based on Native layer is proposed.The method of generating control flow graph of mobile terminal software is given.The generation algorithm of malicious code base and the malicious detection algorithm of software to be tested are designed.The experimental results show that the proposed method can detect malicious code in the Native layer,and has higher detection accuracy compared with similar algorithms.The design of mobile terminal malware detection system is completed,and the malware detection system is preliminarily implemented.The system combines the Native layer detection with the Java layer detection.It not only has the malware detection function,but also compiles the smart contract to realize the sharing of malware information among different organizations.Finally,the test experiment is designed and completed.The experimental results show that the proposed system has higher detection accuracy and faster detection speed.