A Improved Network Risk Assessment Model Based on HMM

Abstract

Abstract-In HMM risk assessment of network, relationship between the nodes is the key part to compute the state transition matrix. The change of IDS alarm level has property of randomness. To calculate the value of the IDS alarm as the input of HMM is based on the Bayesian algorithm model. Using HMM model, probability of attack successful is computed when length of attack sequence is varied. The experimental evidences show new method is veracity and validity.