A Cognitive Model of Immune System for Increasing Security in Distributed Systems

With developing of systems, the security of distributed systems such as Grid is going to be a fundamental challenge. However various methods and tools were used to increasing the security, but the most of them impose a centralized management. In this paper, a cognitive model based on Biological Immune System modeling will be proposed. In the proposed model four groups of autonomous agents are used. Each agent has learning ability and could interact with other agents. The structure of agents is based on Biological Agents that has memory and could use previous experiments. Each group of agents is designed in two levels. In the first level, the model of the agent, its components and the relation between them are presented. In the second level, the interactions of agents are designed and illustrated. The two levels are based on the B cell and T cell lymphocytes modeling. Furthermore, the behavior flowchart of agents will be presented and explained. The proposed model is based on collaboration of the agents and doesn't need centralized management.