The Study Of Assessment And Prediction Methods For Network Security Situation

With the rapid development of information around the world, people more and more rely on network, network security problem also gradually get people’s attention. Due to the fuzziness and randomness of the network system, the traditional single defense can not meet the current needs. Network security situation assessment and prediction provides a new technique for ensuring the safe and stable operation of the network, at the same time, the researchers have deeply studied. The problem of network security situation has not formed the unified optimal solution in the framework and the algorithm, so many new algorithms have been proposed and researchers still need to continue to explore. This paper mainly studies algorithm optimization, the main work is as follows.1. Statistics and analysis of the data of network situation, analyzing the present situation of network security situation at home and abroad to enrich the factors of affecting network security situation and preprocessing the data, so as to improve the speed and efficiency of the training model.2. On the basis of deep understanding of the cloud model and the theory of entropy, network security risk assessment model based on multidimensional cloud model and the entropy weight theory is proposed in this paper. Multidimensional cloud model established the relationship between the qualitative concept and quantitative values, and using the entropy theory to get the weight of each index objectively. Through the digging of the historical data of the network security situation to look for the law, combining with the current network status, and implementing the evaluation of the current network security situation. Thereby, it can improve the objectivity and understandability of the evaluation results.3. In this paper, using particle swarm optimization algorithm to optimize RBF neural network to realize network security situation prediction in the future. By particle swarm optimization algorithm to optimize weights of RBF neural network, it can quickly find the global optimal solution of weights, so as to improve the training speed of the RBF neural network, it can be able to quickly and accurately predict network security situation in the future. In this paper, through the analysis of historical data, and training the PSO- RBF neural network, so that it can predict network security situation in the future quickly and effectively by the PSO-RBF neural network prediction model. Experimental results show that the prediction model can predict the future network security quickly and effectively.