A Computational Model and Multi-Agent Simulation for Information Assurance
Michael VanPutte · Defense Technical Information Center (DTIC) · 2002
The field of information assurance (IA) is too complex for current modeling tools, While security analysts may understand individual mechanisms at a particular moment, the interactions among the mechanisms, combined with evolving nature of the components, make understanding the entire system nearly impossible. This dissertation introduces a computational model of IA called the Social-Technical Information Assurance Model (STIAM). STIAM models organizations, information infrastructures, and human actors as a complex adaptive system. STIAM provides a structured approach to express organizational IA issues and a graphical notation for depicting the elements and interactions. The model can be implemented in a computational system to discover possible adaptive behavior in an IA environment. A multi-agent simulation is presented that introduces several innovations in multi-agent systems including iconnectors, a biologically inspired visual language and mechanism for inter-agent communications. The computational model and simulation demonstrate how complex societies of autonomous entities interact. STIAM can be implemented as a hypothesis generator for scenario development in computer network defensive mechanisms.