Home // International Journal On Advances in Software, volume 7, numbers 1 and 2, 2014 // View article
Authors:
Thomas Zefferer
Peter Teufl
David Derler
Klaus Potzmader
Alexander Oprisnik
Hubert Gasparitz
Andrea Höller
Keywords: Android; power consumption; application classification; malware detection; machine learning
Abstract:
Smartphones and related mobile end-user devices represent key components of mobile computing based solutions and enable end users to conveniently access services and information virtually everywhere and any time. Due to their continuously growing importance and popularity, mobile devices have recently become a common target for malware. Unfortunately, capabilities of malware-detection applications on smartphones are limited, as integrated security features of smartphone platforms such as sandboxing or fine-grained permission models restrict capabilities of third-party applications. These restrictions prevent malware-detection applications from accessing required information for the identification of malware. This renders the implementation of reliable malware-detection solutions on smartphones difficult. To overcome this problem, we propose an alternative malware-detection method for smartphones that relies on the smartphone's measured power consumption. We show that information contained in the measured power consumption of smartphones can in principle be used to identify certain kinds of malware by means of simple threshold-based approaches. We also propose two different machine-learning techniques that allow for a classification of applications according to their power consumption in situations, where disturbing influences prevent an application of simple threshold-based approaches. The capabilities of all proposed techniques have been assessed by means of an evaluation with real-world applications running on physical smartphones. The results of this evaluation process demonstrate the applicability of power consumption based classification and malware-detection approaches in general and of the two proposed machine-learning techniques in particular.
Pages: 150 to 160
Copyright: Copyright (c) to authors, 2014. Used with permission.
Publication date: June 30, 2014
Published in: journal
ISSN: 1942-2628