A Danger Theory Inspired Multi-agent Fusion Model for Network Security Assessment

Abstract

This paper presents a Danger Theory inspired multi-agent fusion model. Danger Theory is a relative new research subject in Artificial Immune System (AIS). It presents new viewpoint and explains some problems in traditional immune theory. This paper describes the principle of Danger Theory, analyzes the differences between it and traditional theory, and also constructs a multi-agent fusion model for network security assessment. It apply the concepts of Danger Theory to fusion model, such as danger signal, danger zone and Antigen Presenting Cells (APC), which are benefit for multi-agent fusion model. An example is presented to illustrate the operation process of the fusion model.