Research In Intrusion Detection Mechanism Of Mobile Intelligent Terminals Based On Cloud Model

With the Internet and the maturity of 3G, 4G technology and development, independent of the operating system convenient mobile intelligent terminal as its convenience and mobility strong characteristics of the rapid development; Users will be intelligent terminal used in all aspects of daily life, such as chatting, shopping, video conference, speech recognition and positioning service, mobile payment, etc. People are enjoying their mobile intelligent terminal bring us convenient, at the same time there are plenty of danger that comes, more and more safety problems points to the mobile intelligent terminal equipment. At present, the biggest threat to intelligent terminal is malicious applications. Mainly including Malicious deduction, rogue theft, privacy to steal, remote control,charges consumption security threats. As the intelligent terminal has many operating system types, each type has the different of security and security standards. In this paper, in-depth study of the current security threat to the biggest Android operating system. Because of a shortage of intelligent terminal system resources and the advantage of the cloud, the paper raises research in intrusion detection mechanism of mobile intelligent terminals based on cloud model.The main work:(1) Because IMACM does not consider after the application is unloaded,it hasn’t been updated by the application applicationed the permissions and caused elevated privileges Communication connection diagram. This paper raises a dynamic tracking that based on weighted directed graph model.The data in the data flow is in the form of weighted directed graph saved in the database. When the application is unloaded, Through reverse breadth and depth first search algorithm find the permissions ascending node and remove it, improve the accuracy of detection.(2)Because the state of t+1moment only relate to the state of t moment and the state does not relate to the dwell time in the HMM.But malicious software is the combination of three permission and attacks time is short, possible attacks have taken place in the current moment and the next moment is normal behavior.In the HMM-2 the state of t+1moment is not only relate to the state of t moment,but also relate to the state of t-1 moment.In the HSMM the probability of dwell time determine the arbitrary distribution state.So HMM-2 combined with HSMM for dectecting the invasion of the cloud.The experiment proved that the proposed algorithm in the Android operating system environment can effectively detect the malicious software applications. And with HMM compared detection method has higher accuracy and lower the rate of false positives. But the complex algorithm to calculate the average test time is slightly higher.