A Botnet Detection Method Based on SCBRNN

With the rapid development of the social network and Internet of things, the complex network environment has led to more serious network security issues. Botnets have always been one of the most important issues in network security. The continuous update of botnet technology has severely influence the network operation of Internet service providers, posing a huge threat to security. Effective detection of botnets is the focus of related security solutions. In the new environment, traditional solutions have become inefficient. In recent years, botnet detection results based on machine learning technology continue to emerge. From the perspective of small batch gradient sample collection, this article optimizes the two-way neural network model and adopts approximate entropy to determine the abnormality of the data, thereby effectively detecting botnets. Research data shows that the model has good performance and can accurately identify botnets. Compared with the traditional model method, when the small batch sampling range is reduced, the accuracy is significantly improved, which provides effective help for Internet service providers to accurately detect botnets, improves service security mechanisms, and improves core competition force.