A Cloud-based Malware Detection Framework

Abstract

Malwares are increasing rapidly. The nature of distribution and effects of malwares attacking several applications requires a real-time response. Therefore, a high performance detection platform is required. In this paper, Hadoop is utilized to perform static binary search and detection for malwares and viruses in portable executable files deployed mainly on the cloud. The paper presents an approach used to map the portable executable files to Hadoop compatible files. The Boyer–Moore-Horspool Search algorithm is modified to benefit from the distribution of Hadoop. The performance of the proposed model is evaluated using a standard virus database and the system is found to outperform similar platforms.