Design and Implementation of an Network Traffic Classification System Based on Machine Learning

Malicious attacks against the Internet are emerging in endlessly. The traditional network architecture makes it impossible for network managers to obtain a complete network view, resulting in difficulties in network security maintenance and management. We have conducted research on network traffic identification, classification, and anomaly detection for network security control, aiming to further improve the efficiency of identification and detection, and enhance the network’s security protection capabilities. A anomaly traffic detection algorithm based on the combination of 1DCNN and LSTM is proposed, which starts from the temporal characteristics of network data and fully utilizes the spatial feature learning ability of CNN and the temporal feature learning ability of LSTM. The experiment shows that the 1DCNN-LSTM algorithm model proposed in this article can effectively reduce the size of the data to be processed, and the accuracy can reach 99.051% on the NSL-KDD data-set, which is about 20% and 10% higher than the accuracy of using CNN or LSTM models alone.