Research On Android Mobile Phone Platform For Anomaly Intrusion Detection

Since 2007, the development of smart mobile phone is more and more quick. They are no longer a smart phone with only basic functions such as calls and messages, but a multi-functional mobile phone with the set of multimedia, office, entertainment and Internet. It can connect to the different remote Internet servers through the mobile Internet, and then downloads large amounts of information stored in the phone. But this information is safe or not, many users do not know. Once the user has downloaded the malicious software or other error messages, it is possible that all their normal information is leaked, which may lead to the immeasurable consequences for users. Therefore, the research of intelligent mobile phone intrusion detection is imminent. According to the status, the paper makes research on smart phone for anomaly intrusion detection.Intrusion detection technology is the key technology in the field of computer security; its development theory has gradually become mature. Due to the late start of smart phone, the research of smart phone in the field of intrusion detection can be said that it’s in the initial stage. This paper, based on the Android smart phone platform, makes the choice of anomaly intrusion detection as the research direction, and combines anomaly intrusion detection with smart phone, so as to enhance the ability to prevent the intrusion of the mobile phone system.The paper, based on the ideas and principles of anomaly intrusion detection, designs and presents a general framework for anomaly intrusion detection system based on the Android smart phone platform, and analyzes the selection of the data feature, the data acquisition module, database model and processing algorithms and response module in detail. It also gives a detailed process of data collection and key code. This paper collects the smart phone operating system information, such as CPU utilization, memory information, process information, network traffic information, and user operating information, then makes the data collected as the data source. After sorting the formation of the training set and test set, the paper uses different SVM algorithm to make experiment and analyze, also records the different results.This experiment in this paper uses the original C-SVM algorithm and improved VC-SVM algorithm to test the data, which is collected by designed systems after compiled. During the experiment, this paper chose the different kernel function and parameter to analyze and calculate the data, and achieved good results. Finally, the results obtained were analyzed and compared in detail. And the results show that, the algorithm is well suited for Android platform, get a high detection rate and low false alarm rate. Combined with the Android smart mobile phone platform, can effectively improve the phone’s active defensive ability, so as to enhance the ability to prevent the intrusion of mobile phone.