Research On Technologies Of Network Survivability Situational Awareness

With the increasing development of network technology, we are facing more large-scale and complex network, network intrusion attacks have can hardly be avoided. The existing network detection and defense technology has been unable to meet the requirements of the development of the network. And now people's attention more focused on the network system in the case of damage, the network system can continue to provide users with key services to meet the needs of the user, which is the survivability of the network, and has became a new direction of network security research.The research of network survivability situation awareness technology mainly includes the network survivability factor extraction, assessment of survivability and predictive technology of survivability. Survivability situational awareness is emphasized from the overall perspective to examine the survivability of the network, which is based on historical data to establish a reasonable and effective forecasting model, and to make effective forecast of the health status of the network system in the future, so as to take corresponding measures to ensure the normal operation of key services to ensure and meet user needs.In this paper, a comprehensive evaluation model based on Improved TOPSIS-Grey Relational Analysis-Entropy Difference (ITOPSIS-GCA-ED) is proposed. Firstly, the fuzzy analytic hierarchy process is used to determine the weight of the key service and its influence factors, avoid the subjective judgment of the weight value. Then use the improved TOPSIS method to make decision on the key service conditions, avoiding the general TOPSIS method to increase or decrease the decision, the positive and negative ideal point is easy to produce reverse problem, and then use the relation analysis to determine the relationship between positive and negative ideal solution and key services, determine the best affiliate degree, and then use the network entropy difference to evaluate the system's key services, and finally get the whole system survivability evaluation and qualitative classification. Through two sets of experimental data, a detailed illustrate the evaluation process of the model and carries on the contrast analysis, show the rationality and validity of the model.In situation forecast, ARIMA(4,1,6) situation forecast model is put forward. Firstly, using the run test method to judge stability of sample data, if the sample data is not smooth, the need for differential treatment until smooth. Then observe sequence of the autocorrelation and partial autocorrelation diagram after the stationary processing, to preliminary judgment and order model, after determining the model, that can be used as a predictive model by white noise test. The experimental results show that the model can reasonably predict the network survivability in the future, and the accuracy of the model is verified.