Unknown Binary Protocol Recognition Algorithm Based on One Class of Classification and One-Dimensional CNN

Table 1

Comparison of advantages and disadvantages of different classification algorithms.

Classification algorithms	Advantages	Deficiency
Bayesian	It has a good effect on uncertain knowledge representation and reasoning, and needs to estimate few parameters, is not sensitive to missing data, and the complexity of time and space is relatively small, so it is easy to expand	A strong conditional independence assumption is needed, which is often not true in practice, so its classification accuracy will decline
Classification algorithm based on association rules	High classification accuracy and strong adaptability	There are still many problems in the execution efficiency, the quality of pruning, and the understandability of the classifier
Artificial neural network	The parallel distributed processing ability is strong, the classification accuracy is high, the distributed storage and learning ability is strong, and it has strong robustness and fault tolerance to noise nerves, especially to nonlinear problems	It requires a large number of parameters, the learning time is long, and the learning process is a black box problem. The output results are difficult to explain, which will affect the confidence of the results
Rough set theory	Can effectively deal with and analyze a variety of incomplete information, and find hidden laws and information	It is sensitive to noise and has low accuracy, so it is usually used in combination with other methods