Parallel-CNN network for malware detection

Abstract

Nowadays, computers and the Internet have become an inseparable part of our life. We accomplish a wide range of our daily tasks through the Internet. A massive number of malwares have been designed annually to infiltrate computers and other electronic devices that endanger their security strikingly. Hence, developing a method that is capable of proactively detect and prevent malware is a perpetual demand. Recently, diverse approaches have been introduced for detecting malware by the help of high-level features and machine learning techniques. Although these methods provide reasonable results, in most of them identifying and extracting proper features from files is one of the most challenging steps. Deep learning techniques that have recently been applied in the area of malware detection, automate the feature extraction operations and represent much better results with respect to multi-layer training. In this study, a novel method is proposed for malware detection by employing a parallel architecture of convolutional neural network (CNN). The proposed method utilises raw bytes of executable files and eliminates the need to extract high-level features. The results of experiments show that the proposed approach can achieve high detection rate, outperforming traditional machine learning based methods which reveals the merit of deep learning techniques in malware detection.