Nearest cluster-based compound network protocol identification

Accurate network protocol identification is the keystone of numerous network activities. Network protocol identification is essential for network and system security in the complex environments such as cloud computing, distributed computing based environment. The most of the network identification classification methods aim to take the advantages of flow statistical features and machine learning techniques. The protocol identification performance is severely affected by limited supervised information and unknown applications. A new method to tackle the problem of unknown applications in the crucial situation of a small supervised training set to achieve effective network protocol identification is implemented in this paper. This method possesses the superior capability of detecting unknown protocols used by unknown applications and utilizing the correlation information among real-world network traffic to boost the identification performance. A theoretical analysis is provided to confirm performance benefit of the method. The comprehensive performance evaluation conducted on real-world network traffic shows that the presented scheme provides proper identification of protocol in the critical network environment. This paper discusses the network protocol identification methodologies. It also includes the details about the implemented proposed work for network protocol identification and unknown protocol detection. The performance of the implemented work is evaluated on the basis of network protocol identification accuracy.