Presenting a Malware Detection System by Implementing Hardware Counters Based on the Multi-Layer Perceptron Neural Network (MLP) and the Dragonfly Optimization Algorithm

Today, one of the most important challenges of information security and communication networks is the increasing number of malware and, consequently, finding suitable ways to protect systems against them. Knowing in time and finding ways to deal with the malicious effects of malware is one of the most important challenges for programmers and information security professionals. Is. Intelligent malware detection systems are able to model malicious behavior well. Extracting appropriate features and using efficient classifiers can improve the performance of such systems. In this paper, a new approach to malware detection is proposed using synergy of the features of the hardware counters and the optimization of the multilayer perceptron neural network classifier. The proposed system is able to identify healthy files from malware by extracting features with high discrimination and also using the neural network optimized by the dragonfly algorithm. In order to evaluate the proposed system, a data set including 168 healthy samples and 437 samples infected with malware is used. The results of the simulations show the higher performance of the proposed category compared to other categories, so that the proposed system has been able to detect the presence of malware-infected files with 86% accuracy.