Research On Intelligent Reverse Analysis Method Of Network Unknown Protocol Syntax

In recent years,with the rapid development of Internet and 5G technology,the frequency of network security news is increasing.Network security is not only related to personal interests,but also related to social stability.So far,the common protocol analysis methods mainly focus on the upper layer protocol,there are many types of unknown protocols,especially private protocols or unpublished protocols,which are difficult to analyze.It is difficult to achieve good results by using common analysis methods.In view of the above situation,this thesis takes the network protocol as the research object,takes the mixed classification of multiple protocols as the goal,and focuses on the intelligent reverse analysis method of unknown protocol syntax.Then it discusses how to optimize the feature extraction algorithm of unknown protocol to make the bitstream protocol feature extraction more accurate and efficient.This thesis mainly includes the following aspects:(1)A new optimization algorithm of unknown protocol feature extraction is designed.By traversing the data,the longest data satisfying the support threshold is found,and the one-dimensional vector is initialized with the length of the longest data.By traversing the dataset to find out the repeat times of each position element,the vector is segmented according to the support.The bad characters are removed to form the initial frequent itemset.Then the noise is removed by support calculation and string statistics,so that pruning can reduce the subsequent workload.Based on this,the feature set is removed and the results are stored in the feature set.Finally,support filtering and de duplication of feature set are performed.The results are saved into the protocol feature library to provide the basis for the follow-up work.(2)An intelligent protocol reverse analysis method is proposed.First,the protocol is transformed into an image,and the data is tagged.Then a convolution neural network model is established.By setting the volume layer,pool layer and full connection layer,the data is mapped to the N class protocol results.Finally,the model is used to test the test data,and then the parameters of the model are adjusted to optimize the model as far as possible.An intelligent protocol analysis method is designed to improve the efficiency of protocol category identification by 15%.(3)Design and implement unknown protocol parsing system.The protocol feature library and intelligent reverse analysis algorithm are used for comprehensive verification,and the evaluation index is designed.The results show that the intelligent protocol syntax reverse analysis method,the convolution neural network algorithm and the unknown protocol feature extraction optimization method can reduce the algorithm running time and improve the protocol recognition accuracy by 5%.By testing the unknown protocol data,it can be shown that the constructed system reduces the response time of the system on the basis of ensuring the accuracy,and provides technical means for network data transmission data flow security prevention and control,data analysis,and accurate detection.