Network-based malware detection and classification by wavelet transformation and machine learning
Mani Arora, Anuj Kumar Gupta · Computational Methods in Science and Technology · 2024
In this era of the internet, malware poses a severe threat to cyber security and modern communication networks. It is very difficult to get protected from the exponential increase of malware threats and its variants with standard security solutions. There are various forms of malware and their variants that can easily compromise the network. Malware analysis, detection, and classification are a real-world tough problem for antivirus manufacturers, as they just rely upon signature-based solutions. Thus, it is really important for detecting and classifying the various variants of malware that affect the network and computer system. The proposed approach uses textural features obtained from image-based malware samples using the wavelet decomposition method. Wavelet decomposition methods for analysing malware using wavelets, extracts the textural features obtained from the image of any malware sample. The feature vectors obtained from the proposed approach reduces the computational complexity and provides a malware detection rate of 98.88% on the benchmark dataset. The wavelet transformation approach for textural feature extraction of malware could be used in the detection of malware attacks in the network. It gives detection accuracy with a relatively small feature vector size. The efficacy and efficiency of the given experimental results depict 98.88% detection accuracy.