2012 5th International Conference on New Technologies, Mobility and Security

Abstract

Anti-malware programs (AM) are well-known powerful security tools that protects against malicious activities. Behaviour-based AMs are seen as suitable candidates to detect unknown malwares in mobile devices. Usually, AM offers different levels of security namely high, medium, and low. The low level offers minimal basic checks in case of low probability of attacks while the high level offers intensive checks in situations of high probability of attacks. Despite their efficiency, behaviour-based anti-malwares are still considered as resource consumer especially in high detection level which reduces considerably the lifetime of handheld devices. In addition, current anti-malwares are not smart to decide when to shift across those levels to reduce both resources consumption and false alarms without scarifying the security level required. To alleviate this problem, we propose a smart anti-malware that can shift between different security levels according to the assets value and the battery status of the resource-constrained device. Such decision is made possible using a nonzero-sum non-cooperative game theoretical model.