Method for behavior-prediction of APT attack based on dynamic Bayesian game
Yang Haopu · 2016
Advanced Persistent Threat, one of the most popular network attacks, has drawn great attention of all over the world because of its huge perniciousness. As the excellent imperceptibility as well as long-term persistence in APT attack, this paper proposes a prediction model based on dynamin Bayesian game. Focusing on the particular features in APT, a corresponding quantitative method is proposed for calculating the behavior payoff. Then, the dynamic Bayesian model is established on the basis of the attack process. The game equilibrium is calculated through the designed solution, which can be used to guide the prediction of APT attack behavior. The experimental result shows that the proposed method achieves great accuracy and effectiveness.