A self-adaptive network traffic classification system with unknown flow detection

Traffic classification technique is an essential tool for network management and system security in terms of high accuracy and fast classification speed. However, in the complex environments such as cloud computing environment, the traditional traffic classifiers will fail facing the unknown protocols that are never seen during training. The state-of-the-art traffic classification methods aim to utilize the semi-supervised learning algorithm to solve the problem. In this paper, we develop a self-adaptive traffic classification system based on semi-supervised learning, which adopts the method, dynamically adding the centers and iterative semi-supervised k-means to choose optimal system parameter and achieve high accuracy. Experiment results derived from real-world traffic are presented to show the effectiveness of the system.