Title: Malicious behaviour identification for Android based on an RBF neural network

Authors: Tianwei Chen; Yongjun Qiang; Chunming Tang; Zairong Wang; Ping Luo; Lingxi Peng

Addresses: School of Information Engineering, Urban Vocational College of Sichuan, Chengdu, 610110, China ' Information and Network Management Office, Sichuan Normal University, Chengdu 610068, China ' College of Mathematics, Guangzhou University, Guangzhou, 510006, China ' Data Recovery Key Laboratory of Sichuan Province, School of Computer Science, Neijiang Normal University, Sichuan, 641100, China ' College of Economics and Statistics, Guangzhou University, Guangzhou, 510006 China ' College of Mechanical and Electrical Engineering, Guangzhou University, Guangzhou, 510006, China

Abstract: A malicious behaviour identification approach for Android that is inspired by radial basis function neural networks is proposed. First, this method takes samples of the malicious behaviour on Android, extracts behaviour features, and integrates data, enabling the use of a radial basis function neural network for identification. Second, the characteristics of the radial basis function neural network's local approximation are used to improve the learning speed, which enhances the quality of the output result. Next, the sum of squares of the distance from each sample in all feature sets to the centre of feature sets is obtained by the k-means clustering algorithm; the minimum value of the distance is used to calculate the weights of the hidden layer node to the output layer node with the least squares recursive method. Further, when the radial basis function neural network training is completed, accurate identification of the malicious behaviour is realised. Finally, the experimental results fully demonstrate that our method improves the accuracy and efficiency of malicious behaviour recognition for Android.

Keywords: RBF neural network; Android malicious behaviour identification; feature sets; local approximation.

DOI: 10.1504/IJSN.2020.109706

International Journal of Security and Networks, 2020 Vol.15 No.3, pp.148 - 154

Received: 10 Oct 2019
Accepted: 30 Oct 2019

Published online: 11 Sep 2020 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article