Research on P2P Botnet Traffic Identification Technology Based on Neural Network

Abstract

The identification of botnet traffic is an important link in the network security. With the continuous improvement of the traditional botnet network traffic identification technology, the identification of the P2P botnet traffic becomes the hot point of the research. The P2P network is a decentralized network model, and the nodes in the network can act as the requesters of the network service, and can respond to the requests of other computers to provide the resources, services and contents, so the difficulty of the malicious traffic identification of the P2P network is greatly increased. The machine learning technology has a wide application in the field of botnet flow identification, but the artificial feature extraction becomes more and more difficult with the change of the shape of the botnet and the control mechanism of the command. To this end, a method for identifying the network traffic of the P2P botnet based on the ResNet convolution neural network is proposed. The experimental results show that the model has good performance and can accurately identify the botnet traffic.