Conceptual Trusted Incident Reaction Architecture

Enterprise networks are continuously growing up and rising connections with various softwares and systems. Their components' security is a tremendous challenge especially due to their heterogeneity and distributed structure. Mechanisms, such as the intrusion detection system, are developed to monitor the security level of those components, their exposure to external attacks or internal failure, and their compliance to target trust level. Although the concept of trust exists for a long time in the computer sciences, it is mainly deployed in the arena of peer-to-peer networking and in specific domains like the e-commerce. The paper proposes a conceptual trusted incident-reaction architecture elaborated firstly based on a multi-agent system that offers the ability to be dynamic and flexible, and secondly based on a decision mechanism that supports the choice of components based on contextual attributes and based on information weighted with trust value.