Threat analysis model of an agent-based vulnerability mitigation mechanism using Bayesian Belief Networks

In this paper we propose a threat analysis model based on Bayesian Belief Networks to analyze and quantify threats towards an Agent-Based Host Enumeration and Vulnerability Scanning Mechanism. Based on several threat scenarios, the model also determines the risks posed towards the mechanism's duty of assessment within the enterprise network. The model construct threats scenarios based on sequenced structure and also forms a multi-scenario threat BBN, where malicious behaviors are interdependent and threat performance aspects are adjusted using network and mechanism specific parameters. Based on predetermined event probabilities within these interdependent structures, threat scenarios likelihoods and risks are driven.