DOI: 10.17587/prin.7.456-463

Methods of Detection of Malicious Software in Operating Systems for Mobile Devices (for Example, the Android Operating System)

S. V. Zhernakov, zhsviit@mail.ru, G. N. Gavrilov, grigrorijgavrilov@mail.ru, Ufa State Aviation Technical University, Ufa, 450008, Russian Federation

The results of the analysis of the Android operating system security for mobile devices and the formalization of malware samples in order to identify features inherent in their behavior are presented. Based on the received results, an sample set was developed, describing the behavior of two types of programs: malicious and safe; the best method for classifying this sample was chosen by means of experiments with using different classification methods (classical, neural networks, and support vector machines). The task of improving the efficiency of malware detection was solved using a technique developed for this purpose based on support vector machines and fuzzy logic. This technique is implemented as a research prototype malware detection system.

Keywords: Android, malware, neural networks, support vector machine, odd logic, technique, operating system

pp. 456–463

For citation:
Zhernakov S. V., Gavrilov G. N. Methods of Detection of Malicious Software in Operating Systems for Mobile Devices (for Example, the Android Operating System), Programmnaya Ingeneria, 2016, vol. 7, no. 10, pp. 456—463.