Botnet Detection Using Recurrent Variational Autoencoder
Botnets are increasingly used by malicious actors, creating increasing threat to a large number of internet users. To address this growing danger, we propose to study methods to detect botnets, especially those that are hard to capture with the commonly used methods, such as the signature based ones and the existing anomaly-based ones. More specifically, we propose a novel machine learning based method, named Recurrent Variational Autoencoder (RVAE), for detecting botnets through sequential characteristics of network traffic flow data including attacks by botnets. We validate robustness of our method with the CTU-13 dataset, where we have chosen the testing dataset to have different types of botnets than those of training dataset. Tests show that RVAE is able to detect botnets with the same accuracy as the best known results published in literature. In addition, we propose an approach to assign anomaly score based on probability distributions, which allows us to detect botnets in streaming mode as the new networking statistics becomes available. This on-line detection capability would enable real-time detection of unknown botnets.