Abstract:Edge computing is a feasible solution for effectively collecting and processing data in industrial Internet of Things (IIoT) systems, and edge security is an important guarantee for edge computing. Fast and accurate classification of malicious code in the whole lift cycle of edge computing is of great significance, which can effectively prevent malicious code from attacking wireless sensor networks and ensure the stable and secure transmission of data in smart devices. Considering that there is a large amount of code reuse in the same malicious code family, making their visual feature similar, many studies use visualization technology to assist malicious code classification. However, traditional malicious code visual classification schemes have the problems such as single image source, weak ability of deep-level feature extraction, and lack of attention to key image details. Therefore, an innovative malicious code visual classification method based on a deep residual network and hybrid attention mechanism for edge security is proposed in this study. Firstly, the malicious code visualization scheme integrates the bytecode file and assembly file of the malware and converts them into a four-channel RGBA image to fully represent malicious code feature information without increasing the computational complexity. Secondly, a hybrid attention mechanism is introduced into the deep residual network to construct an effective classification model, which extracts image texture features of malicious code from two dimensions of the channel and spatial to improve the classification performance. Finally, the experimental results on the BIG2015 and Malimg datasets show that the proposed scheme is feasible and effective and can be widely applied used in various malicious code classification issues, and the classification accuracy rate is relatively higher than the existing better-performing malicious code classification methods.

Fine-grained Classification of Malicious Code Based on CNN and Multi-resolution Feature Fusion

Malicious Code Variant Identification Based on Multiscale Feature Fusion CNNs

Detection and Fine-Grained Classification of Malicious Code Using Convolutional Neural Networks and Swarm Intelligence Algorithms

Malicious Code Detection Based on CNNs and Multi-Objective Algorithm

A Novel Approach to Malicious Code Detection Using CNN-BiLSTM and Feature Fusion

Classification Of Malicious Code Based On Grayscale Texture Fingerprint

Detection of Malicious Code Variants Based on Deep Learning

Feature fusion-based malicious code detection with dual attention mechanism and BiLSTM

Tri-channel Visualised Malicious Code Classification Based on Improved ResNet

Visualization Feature and CNN Based Homology Classification of Malicious Code

Malicious Family Identify Combining Multi-channel Mapping Feature Image and Fine-Tuned CNN

Deep Learning and Regularization Algorithms for Malicious Code Classification

An Efficient Malicious Code Detection System Based on Convolutional Neural Networks

A Malicious Code Detection Method Based on FF-MICNN in the Internet of Things

CNN‐ and GAN‐based classification of malicious code families: A code visualization approach

Malicious Code Classification Method Based on Deep Residual Network and Hybrid Attention Mechanism for Edge Security

Malicious Code Family Classification Method Based on Spatial Pyramid Pooling and Deep Residual Network

IMCFN: Image-based Malware Classification Using Fine-Tuned Convolutional Neural Network Architecture.

Malicious Software Detection Based on Improved Convolution Neural Network

Malware Variants Detection Based on Feature Fusion

Design of Malicious Code Detection System Based on Convolutional Neural Network