Representation of A Modern Strategy for Identifying Bot-Net Attacks in Cloud Computing Using the Neural Network

A bot is a computer infected by malware, controlled remotely by one or more human factors without any user knowledge and will. This research offers a modern strategy for identifying botnet attacks in cloud computing using the neural network (NN). The genetic algorithm has used an improved NN method to detect botnet attacks in this study. In this way, it is noted that the proper characteristics were selected for this. According to historical studies, the features such as Average Flow Size, Packet Average Size, Average Number of Packets, different numbers of flows to the safe destination IP, number of flows to distinctive destination Ports, SYN-SYN/ACK, and Land were chosen. Regression exploration was used to explore the correlation rate between different features and the possibility of botnet attacks. The possibility of botnet attacks is highest for the Average Flow Size. After exploring the correlation of input/output features, Graph Neural Network (GNN), Radial Basis Function (RBF), and Support Vector Machine (SVM) were classified. The best quality of its classification is GNN.