Detection of Botnet Traffic using Deep Learning Approach

Botnets pose a serious threat to network security and have become a significant concern for network administrators. In recent years, machine learning techniques, such as Long Short-Term Memory (LSTM), have shown promise for detecting botnets. The CTU-13 dataset, which contains network traffic data from 13 different botnet scenarios, has become a popular dataset for evaluating the effectiveness of botnet detection methods. However, existing botnet detection systems face several challenges, including the high dimensionality and complexity of network traffic data, the diversity of botnet behaviors, and the presence of noisy and irrelevant features. To overcome these challenges, this project proposes a botnet detection method using LSTM on the CTU-13 dataset. Our objective is to develop a model that can accurately detect botnets in network traffic data. Specifically, we will preprocess the dataset to reduce noise and irrelevant features, train and validate our LSTM model on the preprocessed dataset, and evaluate the performance of our model using standard evaluation metrics. The proposed method achieves a high level of accuracy, outperforming existing botnet detection methods on the CTU-13 dataset. This study can contribute to the development of more effective and efficient botnet detection systems for network security. Overall, this study demonstrates the potential of using LSTM for botnet detection and provides insights into the challenges and opportunities of applying machine learning techniques to network security.