Detection of Viruses on Android Phones Using Multilayer Perceptron Artificial Neural Network

The present study has investigated the analysis and detection of viruses on the Android operating system based on learning machine techniques and data mining. Samples of normal and abnormal programs were collected. Then, static and dynamic analyses were performed on the samples. After dynamic analysis, a number of behavioral characteristics of viruses and normal programs were retrieved and were classified into two categories; one category contained the behavioral characteristics of programs and the other determined whether it was a normal or abnormal program. In the next phase, through the implementation of artificial neural network (ANN) in MATLAB, the ANN learns from the retrieved information. Supervised learning from the retrieved information resulted in the detection of malware in the Android operating system.

Keywords:

detection , Android , malware , artificial neural network , learning machine