Employing The Algorithms Of Random Forest And Neural Networks For The Detection And Analysis Of Malicious Code Of Android Applications

With the rapid development of the Internet, the amount of smart phones has experienced an explosive growth. At the same time, the Android system has occupied the largest share in the operating system market, and it has led to huge economic interests. Because the Android system is open-sourced, it has been chosen as a new attack target by hackers. While Android system brings people great convenience, it has brought serious security problems at the same time. All kinds of viruses keep arising, such as automatically dialing payphone, subscribing spam, and stealing users’accounts and passwords. Facing with the growing Android application library, how to organize, manage and detect it efficiently has become a problem that need to be solved.In this paper, we aim to analyze and detect malicious code of Android applications. The main work is summarized as follows.(1) We made a survey on security mechanism of Android system and the way the malcode attacks. Security mechanism of system includes the security mechanism of Linux kernel itself and the specific security mechanism of Android, such as process sandbox isolation, authority control, and process communication. While analyzing and detecting the Android malware, we summarize the attack mechanism of malcode, such as LKM, Java reflection and polymorphism.(2) We analyzed the function and implementation that the malcode realized on Android system, such as Root crack, code injection, and phone hacking. We further studied Android system based malcode detection method, such as verification detection, proactive defense, and log analysis.(3) We proposed and employed two algorithms, Random Forest and Neuron Netwoks, to analyze and detect Android malcode. Based on the Android application samples as well as the features extracted from the samples provided by our research group. Finally, analysis and detection malicious Android applications using two algorithms. Extensive experimental results show that the detection accuracy achieves over99%which the false alarm rate less than0.5%, and thus demonstrate that this scheme is feasible and effective.This paper focused on security mechanism of Android system itself and the way how malware attack, realized the detection method about Android system malcode which is based on machine learning. Experiments show that, the proposed method can effectively improve the detection rate of Android system malcode.