Search

The identification and classification of network traffic has very strong practical significance. The correct classification of network traffic is the basis of studying the upper network structure. The traditional traffic classification technology mainly relies on port information and load characteristics of data packets. However, in recent years, with the rapid growth of network bandwidth and the gradual complexity of application layer protocols, traditional traffic classification methods have been difficult to meet the needs of the current network environment, which urgently requires the introduction of new theories and technologies to achieve accurate network traffic classification. Combining with the existing network traffic identification and classification methods, this paper studies the network traffic classification method based on machine learning, gives the construction of the classification model based on decision tree, and realizes the function of traffic statistics and classification. Experiments show that the proposed algorithm has higher classification accuracy and is suitable for unbalanced traffic classification compared with many traditional algorithms.

Because the traditional data encryption method remains unchanged after the key is generated, the encryption scheme is single, the encryption cost is high, and it is easy to be cracked. Aiming at the above problems, the dynamic encryption method of user privacy data based on association rules hiding is studied. Based on the DGHV encryption scheme, a dynamic encryption scheme is designed to hide the mined user privacy data association rules. After the key is processed by confounding method, the dynamic encryption process of data is completed. By comparing with the traditional methods, it is verified that the encryption efficiency of the studied method is increased by at least 26.2%, which has the advantage of low encryption cost.

With the penetration of information and communication technology into power production and operation, the dependence of power grid on power communication network becomes stronger and stronger, and the reliability and transmission performance of communication network become the key factors affecting the safe and stable operation of power grid. Based on the analysis and comparison of the existing network traffic models, a hidden Markov model with both long-correlation and short-correlation characteristics is selected to fit and predict the traffic characteristics of the power communication network itself. The article focuses on the development and changes of the production and management of power grid companies, and analyzes the company’s business structure and the bearing relationship of various types of power communication networks. The traffic classification technology of power communication network based on hidden Markov model proposed in this paper has higher accuracy and better stability, and the scores of each evaluation index are higher.