Research on prediction technique of network situation awareness

In this paper we study on the prediction technique of network situation awareness. It has two levels: the high-level situation and the low-level next attack step. The first one includes the indexes and the evaluation results of the network security situation, they are figure form, we use the RBF network to predict them for RBFpsilas self-learning character. Then we use the weighted attack graph to predict the next attack step. The weights represent the probability; the biggest weights indicate the most possible next attack step. The simulations show these prediction methods can offer different prediction capability to satisfy the prediction need of the network situation awareness.