Research On Smartphone Intrusion Detection System

With the development of smartphone in recent years, not only the smartphone has the basic function of call and text message, but also the users can connect the internet directly by it. Meanwhile, more and more viruses made the users in trouble. Much of personal information is stored in the smartphone. It is a great loss to the users if the messages leak out. Therefore, the research on initiative defense of the smartphone is extremely urgent. The research on intrusion detection of the smartphone is put forward.The intrusion detection is used in many fields and develops many products successfully. But the superiority of intrusion detection isn't made full utilized in the smartphone. There are many kinds of technology of intrusion detection. The intrusion detection based on data mining is paid a lot of attentions. The combination of many algorithms in data mining receives a good result. As the lack of the ability on unknown threats of smartphone, the main content in this paper is to use the smartphone as the platform to combine the classification, relevance regulation, clustering and SVM into intrusion detection to improve the initiative defense and to improve the ability of detecting the unknown intrusion.After the proposal, we have done the following work:1) Collect and compute the information of smartphone under normal conditions and attack state, such as the network flow of smart phones, CPU utilization, phone memory and so on.2) Form training set and test sets after get experimental raw data and do the test for the different models, including the DBSCAN algorithm, Bayes algorithm, Apriori algorithm, C-SVC, v-SVC and One-Class-SVM.3) Compute and analyze the experimental data.