Feature identification of unknown protocol

This research gives a brief introduction on the relevant theories and concepts of information theory, which lays a foundation for the application of feature selection algorithm. Afterwards, intensive studies are conducted on the existing protocol identification technologies to analyze their advantages and disadvantages by enumerating relevant algorithms and principles of the pattern matching and data mining. Meanwhile, detailed introduction is also given to the method for analyzing the considerable amount of bit stream data: firstly, relevant theories of data mining are employed to analyze its implication relations, so as to find all the possible candidate strings. Then, the pattern matching algorithm is utilized to make further analysis.