Compound Attack Prediction Method Based on Improved Algorithm of Hidden Markov Model

Network attacks are developing in the direction of concealment, complexity, multi-step, etc., making it difficult to identify and predict. In order to solve the problems such as the difficulty of determining the matching degree of the network attack, the difficulty of predicting the attack intention, and the incorrect calculation of the alarm intent sequence due to the incorrect alarm information, a hidden Markov model based on improved algorithm composite attack prediction is proposed. Firstly, in order to improve the learning ability and adaptability of the algorithm, an improved Baum-Welch algorithm is proposed to train the hidden Markov model (HMM) and generate new HMMs. Then use the Forward algorithm to calculate the HMM with the maximum probability of generating a pre-processed alarm message sequence. When the alarm message sequence is misreported, the attack intent sequence obtained by the classic Viterbi algorithm may be biased. This paper improves the Viterbi algorithm to make the extracted attack intention sequence more accurate. Finally, simulation results show that the model can effectively extract attack intention sequence and improve the accuracy of compound attack prediction.